asking something..

[krz02]

Dear Nathan, i've used your BASNet code but i'm confused.

1. based on my experiment, is it true test data and test result are in different orientation? for example, i put horizontal orientation test data & the result is vertical orientation. sorry if wrong.

2. can you explain what is running_loss, running_tar_loss, ite_num4val and tar ?

thank you before :)

[xuebinqin]

1. Are you using the original basnet_test.py for your testing? I tried that again and the results are correct. Could you please send me a sample of your testing data? Or you can debug the code about "resize" function in the output function of basnet_test.py or the "transpose" and "reshape" function in the dataloader.py. These functions are closely related to the image directions.

2. Since we are using dense supervision, runnning_loss, which denotes the summation of the losses computed from all side outputs, is used for the optimization. running_tar_loss (just for showing not for optimization) denotes the loss computed from the output that used as our final result. The training code suppose to output a model for every certain number (e.g. 2000) of iterations, ite_num4val [1,2000] is the counter for averaging the losses in between two model outputs.

On Wed, Oct 9, 2019 at 5:27 AM evv21 notifications@github.com wrote:

Dear Nathan, i've used your BASNet code but i'm confused.

1.

based on my experiment, is it true test data and test result are in different orientation? for example, i put horizontal orientation test data & the result is vertical orientation. sorry if wrong. 2.

can you explain what is running_loss, running_tar_loss, ite_num4val and tar ?

thank you before :)

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORI5FFSJYXYILA7OATDQNW5Y3A5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HQTMYSQ, or mute the thread https://github.com/notifications/unsubscribe-auth/ADSGORMUQGUP7DX57ACUEFDQNW5Y3ANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[krz02]

Dear Xuebin, thank you for your response

1. Yes, i used original basnet.pth, but i do a minor modification so it can run on my pytorch ( i used pytorch 1.0). the modification only in save_output and inference for each image section (which i added if __name__ == '__main__':freeze_support(). The rest code still same.

2. okay, i understand. in print("[epoch: %3d/%3d, batch: %5d/%5d, ite: %d] train loss: %3f, tar: %3f " is tar an abbrevation? also, in basnet_train, there is no validation function ? because i can't find it.

[krz02]

forget to ask, sorry. if i want to re-train your basnet.pth model with my dataset. can explain how? thankss

[xuebinqin]

1. check our code. The results are the same with yours, in which 3.jpg results a height and width opposite mask. Follows are the shape information of your three images. As you can see, they are the same which means the three images should have the same direction.

   1.jpg (2736, 3648, 3)

   ---

   3.jpg (2736, 3648, 3)

   ---

   2.jpg (2736, 3648, 3)

   But your 3.jpg image does show in a different direction. You could take a look at this link: https://stackoverflow.com/questions/26561800/pil-image-size-returns-the-opposite-width-height It indicates that your 3.jpg may have an EXIF meta data which contains an orientation parameter that tells the image viewers to show in a different direction with the other two images. I guess it won't introduce errors in training or testing steps because skimage or PIL reads them with the exact correct dimension. If you want to them to show correctly, you could google it.

2. yes "tar" is the abbreviation. You can load the pre-trained model first and continue to train it. Add net.load_state_dict(torch.load("path/basnet.pth")) after line 115: net = BASNet(3, 1) in basnet_train.py and then train the network with your own data.

On Wed, Oct 9, 2019 at 11:53 PM evv21 notifications@github.com wrote:

forget to ask, sorry. if i want to re-train your basnet.pth model with my dataset. can explain how? thankss

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORLBWRPLNZMEO3NQVYTQN27OFA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEA22HMQ#issuecomment-540386226, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORMTQQF3TS3HWZVAP2LQN27OFANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[krz02]

1. That's happen because 3.jpg orientation actually is more likely to landscape, then skimage/PIL rotate it to the correct orientation, right ?

2. • Then, "tar" means target ?
   • Alright, i've tried adding net.load_state_dict(torch.load("path/basnet.pth")) and tuning it using my dataset, it looks works. So i assume no need to add following code to basnet_train.py ? because i've read, to do tuning of pre-trained model and reset final fully connected layer need follows code. model_conv = torchvision.models.resnet18(pretrained=True) for param in model_conv.parameters(): param.requires_grad = False num_ftrs = model_conv.fc.in_features model_conv.fc = nn.Linear(num_ftrs, len(class_name)

• Sorry, could you share the evaluation code too for f-measure, relax and MAE ?

sorry for asking many questions because i'm still beginner ..

[xuebinqin]

1.That's happen because 3.jpg orientation actually is more likely to landscape, then skimage/PIL rotate it to the correct orientation, right ? The skimage/PIL didn't rotate the image. They read the image correctly. The EXIF contains the parameter that enables your image viewer and OS to show the rotated one.

2. "tar" denotes "target" which indicate our final target output. If you want to fix the resnet (we are using resnet34 as shown in the line 108 of BASNet.py) pre-trained model parameters in your fine tunning, you could add these codes shown in your email. Otherwise just ignore them. BTW, we didn't use the fully connected layer of the resnet.

On Fri, Oct 11, 2019 at 2:40 AM evv21 notifications@github.com wrote:

1.

That's happen because 3.jpg orientation actually is more likely to landscape, then skimage/PIL rotate it to the correct orientation, right ? 2.

• Sorry, but "tar" denotes ?
• Alright, i've tried adding net.load_state_dict(torch.load("path/basnet.pth")) and tuning it using my dataset, it looks works. So i assume no need to add following code to basnet_train.py ? or still add it? because i've read, to do tuning of pre-trained model and reset final fully connected layer need follows code. model_conv = torchvision.models.resnet18(pretrained=True) for param in model_conv.parameters(): param.requires_grad = False num_ftrs = model_conv.fc.in_features model_conv.fc = nn.Linear(num_ftrs, len(class_name)

sorry for asking many questions ..

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORNDCAWLPQWWYKWIT23QOA3W5A5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEA7JTOI#issuecomment-540973497, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORI37X3R4CC6VTYUFALQOA3W5ANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[krz02]

1. Ohh .. i seee, thank u for your explanation ..

2. Alright then. For now, i just need to re-train the basnet.pth to my dataset. Btw, Could u share the evaluation code too for f measure, relaxed and MAE?

[xuebinqin]

I would like to upload the eval code in python to github. But now they are a bit messy. I am currently work on my PHD thesis and don't have time to organize them. I may upload it later. Sorry for the inconvenience.

On Fri, Oct 11, 2019 at 9:19 AM evv21 notifications@github.com wrote:

1.

Ohh .. i seee, thank u for your explanation .. 2.

Alright then. For now, i just need to re-train the basnet.pth to my dataset. Btw, Could u share the evaluation code too for f measure, relaxed and MAE?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORO54SDCLQDJWW4EGRDQOCKPNA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBAKNVQ#issuecomment-541107926, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORKRPNEYCKRBUL7GH63QOCKPNANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[krz02]

Alright then, thanks for the explanation. Sorry for bothering you. next time, if there's something I don't understand, I'll ask again.

[xuebinqin]

You're very welcome. Thank you for your interests to our project.

On Fri, Oct 11, 2019 at 9:50 AM evv21 notifications@github.com wrote:

Alright then, thanks for the explanation. Sorry for bothering you. next time, if there's something I don't understand, I'll ask again.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORMDNAVQRFZX4ZTD2ODQOCOC5A5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBANHBA#issuecomment-541119364, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORO5MRTXPMLYD36ZYD3QOCOC5ANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[krz02]

You're welcome :). Btw why do not use fully connected layer?

[xuebinqin]

Because our application is pixel-wise image segmentation. We use the convolutional layers for extracting deep features of each pixel. The last fully connected layers are used for image-wise classification, which is unnecessary in our application.

On Fri, Oct 11, 2019 at 9:53 AM evv21 notifications@github.com wrote:

You're welcome :). Btw why do not use fully connected layer?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORKC6MBXOWKDW2JEVELQOCON7A5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBANP2Y#issuecomment-541120491, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORLDIM34RA5OWEJZEQ3QOCON7ANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[krz02]

That means, most of image segmentation task dont use final /last fully connected layer ? & Last fully connected layer only used for task that recognize specific class/ image ?

[xuebinqin]

Yes,

On Fri, Oct 11, 2019 at 11:08 AM evv21 notifications@github.com wrote:

That means, most of image segmentation task dont use final /last fully connected layer ? & Last fully connected layer only used for task that recognize specific class/ image ?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORNEXN4L3NPBL2WR7D3QOCXHPA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBAT6GY#issuecomment-541146907, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORJOCZMFBTOOIUE4QF3QOCXHPANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[krz02]

Sorry for asking again, in basnet.train. is 'validation' method is done ? Because i can't find it ( validate to train dataset) and i read on your paper calculating F beta score, what is the beta value?

[xuebinqin]

sorry, what do you mean about 'validation' method (validate to train dataset)? I guess what you mean about the validation to train dataset is the same with the computing the training loss.

On Fri, Oct 11, 2019 at 9:59 PM evv21 notifications@github.com wrote:

Sorry for asking again, in basnet.train. is 'validation' method is done ? Because i can't find it ( validate to train dataset)

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORIEXCJB56UTYNSKLKTQOFDSXA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBBU3CQ#issuecomment-541281674, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORKLHILIM67FQUR4L3TQOFDSXANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[krz02]

-Ah, yes. I mean validation to train dataset. Sorry. Then, how i determine its overfitting or underfit? Too low training loss can also be interpreted overfitt ?

• also, i want to ask about Fβ score, what is the value of β?

[krz02]

Sorry, missed when reading your paper. value β^2 = 0.3

[xuebinqin]

Ah, yes. I mean validation to train dataset. Sorry. Then, how i determine its overfitting or underfit? Too low training loss can also be interpreted overfitt ?

in this application, the situation is a bit special. We tested on six different public datasets and their distributions are different actually. Using any of the validation set won't make sure that you get the optimal model. According to our many times experiment (we were using MSRA-B as validation set before), we found that pick the model with the smallest training loss usually performs better than that with the smallest validation loss. So in our experiment settings, we choose to find the final model from a few models with the smallest training losses. Validation set was not using then.

also, i want to ask about Fβ score, what is the value of β?

\beta is to weight the precision and recall. please refer to the paper and related references.

On Sun, Oct 13, 2019 at 1:11 PM evv21 notifications@github.com wrote:

-Ah, yes. I mean validation to train dataset. Sorry. Then, how i determine its overfitting or underfit? Too low training loss can also be interpreted overfitt ?

• also, i want to ask about Fβ score, what is the value of β?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORMMTNJ74DWHACTF5DLQONXH3A5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBC5RVI#issuecomment-541448405, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORLDIOXRYBD3J5QF4WTQONXH3ANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[xuebinqin]

BTW, you should check the training curve to determine if that was underfit or not.

On Sun, Oct 13, 2019 at 1:24 PM Xuebin Qin xuebin@ualberta.ca wrote:

Ah, yes. I mean validation to train dataset. Sorry. Then, how i determine its overfitting or underfit? Too low training loss can also be interpreted overfitt ?

in this application, the situation is a bit special. We tested on six different public datasets and their distributions are different actually. Using any of the validation set won't make sure that you get the optimal model. According to our many times experiment (we were using MSRA-B as validation set before), we found that pick the model with the smallest training loss usually performs better than that with the smallest validation loss. So in our experiment settings, we choose to find the final model from a few models with the smallest training losses. Validation set was not using then.

also, i want to ask about Fβ score, what is the value of β?

\beta is to weight the precision and recall. please refer to the paper and related references.

On Sun, Oct 13, 2019 at 1:11 PM evv21 notifications@github.com wrote:

-Ah, yes. I mean validation to train dataset. Sorry. Then, how i determine its overfitting or underfit? Too low training loss can also be interpreted overfitt ?

• also, i want to ask about Fβ score, what is the value of β?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORMMTNJ74DWHACTF5DLQONXH3A5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBC5RVI#issuecomment-541448405, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORLDIOXRYBD3J5QF4WTQONXH3ANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[krz02]

Additional, so when you doing training before using MSRA-B. Did you split the dataset into Training set & Validation set ? Or just fed them all MSRA- B dataset as training image ? Sorry, im a bit confuse. Because when i re-train your basnet.pth, i just use all my dataset as training images.( In my case, i have 100 images per category / class and i have 7 category/ class, so its about 700 images),

Btw, do u have an email? So i can contact u personally ?

[xuebinqin]

Split MSRA-B as training (2500), valid (500) and test set (2000) according to one paper. Sorry, I forget the paper name.

On Sun, Oct 13, 2019 at 2:02 PM evv21 notifications@github.com wrote:

Additional, so when you doing training before using MSRA-B. Did you split the dataset into Training set & Validation set ? Or just fed them all MSRA- B dataset as training image ?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORK6QLS4JOVOMUYYHWLQON5FDA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBC6XWQ#issuecomment-541453274, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORN47XOSHGX7PHKRLDTQON5FDANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[krz02]

How to add the validation set in your code basnet_train ? Or when u used validation set, u used another code (different than basnet_train) to do the validation ?

On Wed, Oct 16, 2019, 08:31 Xuebin Qin notifications@github.com wrote:

Split MSRA-B as training (2500), valid (500) and test set (2000) according to one paper. Sorry, I forget the paper name.

On Sun, Oct 13, 2019 at 2:02 PM evv21 notifications@github.com wrote:

Additional, so when you doing training before using MSRA-B. Did you split the dataset into Training set & Validation set ? Or just fed them all MSRA- B dataset as training image ?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORK6QLS4JOVOMUYYHWLQON5FDA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBC6XWQ#issuecomment-541453274 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/ADSGORN47XOSHGX7PHKRLDTQON5FDANCNFSM4I66AUQQ

.

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=AIJ3MDFOEHMSNIBFJZMW7TLQOZVGPA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBKXIZQ#issuecomment-542471270, or unsubscribe https://github.com/notifications/unsubscribe-auth/AIJ3MDGYINIYCFUKLBXEU3DQOZVGPANCNFSM4I66AUQQ .

[xuebinqin]

For the training dataset we used in our paper, we didn't use validation set as described in our paper. For adding validation in the code, you can add another dataloader for loading your validation dataset. Then you can use your tranined model checkpoint to validate it.

On Wed, Oct 16, 2019 at 5:00 AM evv21 notifications@github.com wrote:

How to add the validation set in your code basnet_train ? Or when u used validation set, u used another code (different than basnet_train) to do the validation ?

On Wed, Oct 16, 2019, 08:31 Xuebin Qin notifications@github.com wrote:

Split MSRA-B as training (2500), valid (500) and test set (2000) according to one paper. Sorry, I forget the paper name.

On Sun, Oct 13, 2019 at 2:02 PM evv21 notifications@github.com wrote:

Additional, so when you doing training before using MSRA-B. Did you split the dataset into Training set & Validation set ? Or just fed them all MSRA- B dataset as training image ?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub <

https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORK6QLS4JOVOMUYYHWLQON5FDA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBC6XWQ#issuecomment-541453274

, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/ADSGORN47XOSHGX7PHKRLDTQON5FDANCNFSM4I66AUQQ

.

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub < https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=AIJ3MDFOEHMSNIBFJZMW7TLQOZVGPA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBKXIZQ#issuecomment-542471270 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AIJ3MDGYINIYCFUKLBXEU3DQOZVGPANCNFSM4I66AUQQ

.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORM7RFHCEJA6T7SE23TQO3X6RA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBMCFRY#issuecomment-542646983, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORNPACF7SMDX3XVR45DQO3X6RANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

[krz02]

Hmm, before you said valid (500) it means validation set right ? Valid(500) same like validation set ? Sorry, because im confused

On Thu, Oct 17, 2019, 02:11 Xuebin Qin notifications@github.com wrote:

For the training dataset we used in our paper, we didn't use validation set as described in our paper. For adding validation in the code, you can add another dataloader for loading your validation dataset. Then you can use your tranined model checkpoint to validate it.

On Wed, Oct 16, 2019 at 5:00 AM evv21 notifications@github.com wrote:

How to add the validation set in your code basnet_train ? Or when u used validation set, u used another code (different than basnet_train) to do the validation ?

On Wed, Oct 16, 2019, 08:31 Xuebin Qin notifications@github.com wrote:

Split MSRA-B as training (2500), valid (500) and test set (2000) according to one paper. Sorry, I forget the paper name.

On Sun, Oct 13, 2019 at 2:02 PM evv21 notifications@github.com wrote:

Additional, so when you doing training before using MSRA-B. Did you split the dataset into Training set & Validation set ? Or just fed them all MSRA- B dataset as training image ?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub <

https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORK6QLS4JOVOMUYYHWLQON5FDA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBC6XWQ#issuecomment-541453274

, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/ADSGORN47XOSHGX7PHKRLDTQON5FDANCNFSM4I66AUQQ

.

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <

https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=AIJ3MDFOEHMSNIBFJZMW7TLQOZVGPA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBKXIZQ#issuecomment-542471270

, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/AIJ3MDGYINIYCFUKLBXEU3DQOZVGPANCNFSM4I66AUQQ

.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORM7RFHCEJA6T7SE23TQO3X6RA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBMCFRY#issuecomment-542646983 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/ADSGORNPACF7SMDX3XVR45DQO3X6RANCNFSM4I66AUQQ

.

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=AIJ3MDEPR7Z2QVDASC7IBBDQO5ROZA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBNTTAA#issuecomment-542849408, or unsubscribe https://github.com/notifications/unsubscribe-auth/AIJ3MDDJGRQNYC74QC6NGSLQO5ROZANCNFSM4I66AUQQ .

[xuebinqin]

Yes, it is. Follows are the split file names of each part.

On Thu, Oct 17, 2019 at 4:57 AM evv21 notifications@github.com wrote:

Hmm, before you said valid (500) it means validation set right ? Valid(500) same like validation set ? Sorry, because im confused

On Thu, Oct 17, 2019, 02:11 Xuebin Qin notifications@github.com wrote:

For the training dataset we used in our paper, we didn't use validation set as described in our paper. For adding validation in the code, you can add another dataloader for loading your validation dataset. Then you can use your tranined model checkpoint to validate it.

On Wed, Oct 16, 2019 at 5:00 AM evv21 notifications@github.com wrote:

How to add the validation set in your code basnet_train ? Or when u used validation set, u used another code (different than basnet_train) to do the validation ?

On Wed, Oct 16, 2019, 08:31 Xuebin Qin notifications@github.com wrote:

Split MSRA-B as training (2500), valid (500) and test set (2000) according to one paper. Sorry, I forget the paper name.

On Sun, Oct 13, 2019 at 2:02 PM evv21 notifications@github.com wrote:

Additional, so when you doing training before using MSRA-B. Did you split the dataset into Training set & Validation set ? Or just fed them all MSRA- B dataset as training image ?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub <

https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORK6QLS4JOVOMUYYHWLQON5FDA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBC6XWQ#issuecomment-541453274

, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/ADSGORN47XOSHGX7PHKRLDTQON5FDANCNFSM4I66AUQQ

.

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <

https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=AIJ3MDFOEHMSNIBFJZMW7TLQOZVGPA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBKXIZQ#issuecomment-542471270

, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/AIJ3MDGYINIYCFUKLBXEU3DQOZVGPANCNFSM4I66AUQQ

.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub <

https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORM7RFHCEJA6T7SE23TQO3X6RA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBMCFRY#issuecomment-542646983

, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/ADSGORNPACF7SMDX3XVR45DQO3X6RANCNFSM4I66AUQQ

.

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub < https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=AIJ3MDEPR7Z2QVDASC7IBBDQO5ROZA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBNTTAA#issuecomment-542849408 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AIJ3MDDJGRQNYC74QC6NGSLQO5ROZANCNFSM4I66AUQQ

.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGOROF26MWQLSCKZJTYFLQPBAJFA5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBPV2EY#issuecomment-543120659, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORJFHOXGKIGEA25JI63QPBAJFANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/

0_0_147.png 0_0_155.png 0_0_280.png 0_0_284.png 0_0_431.png 0_0_547.png 0_0_579.png 0_0_735.png 0_0_77.png 0_0_818.png 0_0_951.png 0_10_10134.png 0_10_10135.png 0_10_10176.png 0_10_10343.png 0_10_10392.png 0_10_10396.png 0_10_10484.png 0_10_10507.png 0_10_10545.png 0_10_10938.png 0_11_11179.png 0_11_11219.png 0_11_11281.png 0_11_11313.png 0_11_11325.png 0_11_11435.png 0_11_11459.png 0_11_11480.png 0_11_11533.png 0_11_11557.png 0_11_11830.png 0_11_11875.png 0_11_11924.png 0_11_11987.png 0_12_12000.png 0_12_12048.png 0_12_12072.png 0_12_12322.png 0_12_12372.png 0_12_12429.png 0_12_12597.png 0_12_12693.png 0_12_12705.png 0_12_12750.png 0_12_12771.png 0_12_12816.png 0_12_12833.png 0_12_12884.png 0_12_12921.png 0_12_12923.png 0_12_12974.png 0_13_13036.png 0_13_13264.png 0_13_13308.png 0_13_13339.png 0_13_13349.png 0_13_13356.png 0_13_13361.png 0_13_13420.png 0_13_13450.png 0_13_13453.png 0_13_13460.png 0_13_13845.png 0_14_14093.png 0_14_14181.png 0_14_14532.png 0_14_14830.png 0_14_14901.png 0_15_15264.png 0_15_15345.png 0_15_15522.png 0_15_15622.png 0_15_15644.png 0_15_15742.png 0_16_16030.png 0_16_16079.png 0_16_16403.png 0_16_16629.png 0_16_16947.png 0_16_16968.png 0_16_16989.png 0_17_17251.png 0_17_17369.png 0_17_17388.png 0_17_17407.png 0_17_17526.png 0_17_17569.png 0_17_17660.png 0_18_18219.png 0_18_18254.png 0_18_18314.png 0_18_18358.png 0_18_18661.png 0_18_18690.png 0_18_18699.png 0_18_18720.png 0_18_18725.png 0_18_18732.png 0_18_18763.png 0_18_18804.png 0_18_18816.png 0_18_18852.png 0_18_18880.png 0_18_18961.png 0_19_19136.png 0_19_19200.png 0_19_19214.png 0_19_19686.png 0_19_19820.png 0_19_19930.png 0_1_1626.png 0_1_1664.png 0_1_1696.png 0_1_1865.png 0_20_20085.png 0_20_20144.png 0_20_20368.png 0_20_20724.png 0_20_20772.png 0_21_21001.png 0_21_21093.png 0_21_21135.png 0_21_21147.png 0_21_21332.png 0_21_21413.png 0_21_21703.png 0_21_21938.png 0_21_21974.png 0_22_22047.png 0_22_22147.png 0_22_22219.png 0_22_22502.png 0_23_23003.png 0_23_23069.png 0_23_23355.png 0_23_23577.png 0_23_23645.png 0_23_23666.png 0_23_23732.png 0_24_24071.png 0_24_24101.png 0_24_24234.png 0_24_24368.png 0_24_24431.png 0_24_24440.png 0_24_24442.png 0_24_24455.png 0_24_24456.png 0_24_24492.png 0_24_24500.png 0_24_24532.png 0_24_24670.png 0_24_24866.png 0_24_24906.png 0_24_24951.png 0_24_24971.png 0_24_24973.png 0_2_2225.png 0_2_2546.png 0_2_2551.png 0_2_2721.png 0_2_2950.png 0_3_3142.png 0_3_3276.png 0_3_3308.png 0_3_3317.png 0_3_3327.png 0_3_3344.png 0_3_3345.png 0_3_3374.png 0_3_3434.png 0_3_3467.png 0_3_3473.png 0_3_3559.png 0_3_3654.png 0_3_3692.png 0_4_4038.png 0_4_4316.png 0_4_4666.png 0_4_4838.png 0_4_4870.png 0_4_4976.png 0_5_5031.png 0_5_5108.png 0_5_5131.png 0_5_5189.png 0_5_5268.png 0_5_5566.png 0_5_5634.png 0_5_5673.png 0_5_5706.png 0_5_5948.png 0_5_5953.png 0_6_6087.png 0_6_6112.png 0_6_6227.png 0_6_6287.png 0_6_6574.png 0_6_6618.png 0_6_6823.png 0_6_6826.png 0_6_6830.png 0_6_6833.png 0_6_6847.png 0_6_6876.png 0_6_6917.png 0_6_6975.png 0_7_7129.png 0_7_7540.png 0_7_7586.png 0_7_7677.png 0_7_7697.png 0_7_7796.png 0_7_7817.png 0_7_7923.png 0_7_7981.png 0_8_8108.png 0_8_8241.png 0_8_8454.png 0_8_8648.png 0_8_8814.png 0_8_8976.png 0_9_9002.png 0_9_9003.png 0_9_9033.png 0_9_9127.png 0_9_9132.png 0_9_9135.png 0_9_9187.png 0_9_9264.png 0_9_9419.png 0_9_9496.png 0_9_9638.png 0_9_9699.png 0_9_9837.png 0_9_9880.png 0_9_9990.png 10_00000006_004.png 10_00000010_010.png 10_00000013_001.png 10_00000023_007.png 10_00000024_023.png 10_00000027_010.png 10_00000033_020.png 10_00000034_013.png 10_00000034_014.png 10_00000043_019.png 10_00000067_018.png 10_00000069_002.png 10_00000094_012.png 10_00000094_022.png 10_00000098_007.png 10_00000098_010.png 10_100814280_70659ccbfa.png 10_114197289_0b92e3abcc.png 10_118134441_c58fc0719f.png 10_132474673_88eaf528f5.png 10_134815623_dcbd7ad8e0.png 10_137385456_396ee799c5.png 10_139375137_1d54fa7e20.png 10_142031788_e469594839.png 10_152109747_5678ebbb1c.png 10_15955708.montblanc1.png 10_162330937_390a32901a.png 10_167944629_67a359c464.png 10_178480338_a73f99f0ba.png 10_184503160_143a47edb1.png 10_206775389_ad06667dc2.png 10_217098809_9fddf98ceb.png 10_229961305_0a965550ec.png 10_23411894_f0827bc59c.png 10_234794066_610aa191d1.png 10_245737491_33814f1975.png 10_245741115_52f74157d9.png 10_247193222_122354a299.png 10_256056095_92cb6f9946.png 10_256246201_af99191b82.png 10_257430398_b468551510.png 10_257476956_dd7b951ab3.png 10_259829970_8ad028f44f.png 10_259830185_a6b7dd03eb.png 10_259871222_13311d0e2f.png 10_260372597_3269d98a53.png 10_260610820_773d74b04a.png 10_261185388_23b72d8b06.png 10_261717707_bb5b8e0b80.png 10_263808439_8c5342f1eb.png 10_265870892_766e28e379.png 10_265917413_40738f7de1.png 10_266085492_354fea0ae0.png 10_266284290_829ababfdb.png 10_266448508_94af676059.png 10_266501049_f8f3228c06.png 10_266505582_b6bea7d285.png 10_266618383_e120335223.png 10_266684747_d96b086291.png 10_266685287_18eb05146e.png 10_266685774_034585ce2d.png 10_266691786_6d0cd9edb2.png 10_266774009_d85fd2d1cb.png 10_266939706_4c0698df47.png 10_266943402_f88493c435.png 10_266985450_6d8e8f730d.png 10_266996498_e98c27d9fa.png 10_266996502_b334ade588.png 10_267014693_4f1ee75c4a.png 10_267183852_ce52f4236d.png 10_267283020_efc5754cec.png 10_267294594_3ac14453d8.png 10_267300600_05f6c6fee6.png 10_267300813_22fcf61089.png 10_267303537_2872e24990.png 10_267303541_20c1da051c.png 10_267311260_f18be257e2.png 10_267393563_c72e472e3c.png 10_267463817_50230632ce.png 10_267466379_f8bbb5fcc8.png 10_267551886_bddb4ac4b6.png 10_267558735_68fd4bd891.png 10_267640333_d68dd4aa2e.png 10_267689043_8ed546e74e.png 10_267689217_5f9b7de27c.png 10_267689464_3dd591c84a.png 10_267690371_4874327dcc.png 10_267693240_8ce8970324.png 10_267693798_74248c315b.png 10_267767008_59aabbd596.png 10_267778234_50c3b2c18d.png 10_267781157_7f33392f45.png 10_267782373_d2cbfb0560.png 10_267782444_2e0050ebd9.png 10_267787440_0399594cbe.png 10_267787613_cd71b51421.png 10_267795560_bcae5fb9db.png 10_267803933_d803fdcaea.png 10_267849716_999316bd5f.png 10_267907104_92b81c94cd.png 10_267920434_fbb13ef558.png 10_267921452_ed232ffb29.png 10_267923236_a7d4c3971f.png 10_267974673_5cd8c01d05.png 10_268001653_da15cc1bb1.png 10_268008679_05d9f1e40f.png 10_268012766_eba8915621.png 10_268012886_472fbedb54.png 10_268023025_784a1ec774.png 10_268032800_bfc9cd4caf.png 10_268037364_97cb639b9e.png 10_268050666_fd8484795a.png 10_268051804_e75c8bd959.png 10_268067924_c212238b2b.png 10_268092274_58b19335ec.png 10_268095200_956217bde8.png 10_268121720_f0d8c64b88.png 10_268131111_8fbe13baef.png 10_268132722_ee1448d776.png 10_268139713_919734cac6.png 10_268142756_1007aab622.png 10_268145469_9c89086dd8.png 10_268146459_503a3645fe.png 10_268150486_bc129aa419.png 10_268167124_b277130727.png 10_268171006_3b22a27896.png 10_268181570_6369b8852f.png 10_268210168_c2c0833ec7.png 10_268211018_3b6c9d60fe.png 10_268213254_f293aaba36.png 10_268213338_b07e31403f.png 10_268214583_21290bd084.png 10_268217148_13cbed044f.png 10_268227714_a9172f0585.png 10_268231974_e051f78f99.png 10_268232243_8022b3ca0a.png 10_268232438_d4c7878070.png 10_268232457_4d016499a5.png 10_268232543_47709ca249.png 10_268233109_174ef0fd9b.png 10_268242883_22851fd122.png 10_268246654_5600cde952.png 10_268256903_3b133255f4.png 10_268257878_a7f3698218.png 10_268260683_84238c97a4.png 10_268262591_b9821c9af8.png 10_268262624_e2c6dfa278.png 10_268263570_c28d1e707b.png 10_268272471_95133f344c.png 10_268283312_da35ac9f4c.png 10_268283436_b55b7f7190.png 10_268283665_4ef68e1d4d.png 10_268283692_95ce1b1fb3.png 10_268284017_749118c309.png 10_268284540_ff074283d2.png 10_268284552_7906464eeb.png 10_268284576_f631047f15.png 10_268284865_e4bc4109fe.png 10_268285156_43e66d3ffd.png 10_268289886_b23a2eb6a4.png 10_268302285_c68ec99f56.png 10_2719989.izfennel_p320001972.png 10_28113685.growl.png 10_28242550.img_5000.png 10_29475162.fuzhou620fupeixi1.png 10_29912710.xcountry.png 10_30169653.kazi.png 10_33152466.catlamp.png 10_38136691.kleinekater.png 10_39452112.img_0488.png 10_39881820.b91u0017.png 10_39898384.img_0501.png 10_40682501.img_1575.png 10_40837312.dsc_0800.png 10_41100185.dsc_5762a.png 10_41491048.streakedspiderhunter.png 10_43047908.nandaihe.png 10_43137038.copyofds20050507_0056awfairliegallop.png 10_43642601.pasqueflowerpasqueflowerearlyeve.png 10_44399460.pinkflower.png 10_44859872.5.png 10_44859874.6.png 10_46084290.commontern2.png 10_46084614.commontern5.png 10_46642384.roadingjuly02006.png 10_46642387.roadingjuly02009.png 10_49930792.initiationjunior1134.png 10_49930794.initiationjunior1140.png 10_49930795.initiationjunior1141.png 10_49930799.initiationjunior1145.png 10_49930800.initiationjunior1146.png 10_49930801.initiationjunior1147.png 10_49930805.initiationjunior1151.png 10_49930806.initiationjunior1152.png 10_52202423.dscn0194.png 10_52202426.dscn0202.png 10_52202429.dscn0205.png 10_52202442.dscn0223.png 10_52202444.dscn0225.png 10_52202446.dscn0227.png 10_52202449.dscn0230.png 10_52202451.dscn0240.png 10_52202452.dscn0241.png 10_52202467.dscn0267.png 10_52202479.dscn0285.png 10_54628934.jan8_06_554.png 10_54628935.jan8_06_555.png 10_54646369.1ethansillykid.png 10_55651852.carlystreetweballey6.png 10_55651903.carlystreetwebstreet9.png 10_55657233.dsc_3867_y_jp.png 10_55768607.bitsy111.png 10_55974006.arabfillycanter.png 10_59271034.127_2729.png 10_59271042.140_4092.png 10_59271063.img_0605.png 10_59271070.img_1436.png 10_59271073.img_1750.png 10_59271089.img_3385.png 10_59271094.img_3692.png 10_59271105.img_9130.png 10_60607759.rosebreastedgrosbeakfemale.png 10_60666278.easterncottontail.png 10_60866226.cedarwaxwing.png 10_60866769.turkeyvulture2.png 10_64229307.utd1927n.png 10_64386031.3qqfbtes.housefinch.png 10_65345059.baxlulsq.img_7128_editsrgb.png 10_66830581.uvuxhte2.sept10_06_701.png 10_66830582.bxg9hxcr.sept10_06_702.png 10_66830583.sdhhngr2.sept10_06_703.png 10_66830587.j2qal6cp.sept10_06_706.png 10_66830631.ag3y7hqi.sept10_06_747.png 10_66830632.ncrvs2io.sept10_06_748.png 10_66830633.6hispsw2.sept10_06_749.png 10_67151572.bn0308dt.png 10_73754853_140eb787dd.png 10_8168705.izcenoadj_p7070026640.png 10_93082635_7f4aa15eed.png 10_95071511_4720e4e540.png 10_97146107_ddc8084759.png 10_97320358_cc615b9019.png 10_99533452_3c0e47e42d.png 1_25_25164.png 1_25_25226.png 1_25_25257.png 1_25_25389.png 1_25_25954.png 1_26_26026.png 1_26_26104.png 1_26_26113.png 1_26_26213.png 1_26_26233.png 1_26_26366.png 1_26_26640.png 1_26_26739.png 1_26_26763.png 1_26_26833.png 1_26_26847.png 1_26_26853.png 1_26_26945.png 1_26_26982.png 1_27_27003.png 1_27_27033.png 1_27_27543.png 1_27_27705.png 1_28_28003.png 1_28_28373.png 1_28_28391.png 1_28_28407.png 1_28_28717.png 1_28_28763.png 1_28_28970.png 1_29_29028.png 1_29_29099.png 1_29_29138.png 1_30_30022.png 1_30_30051.png 1_30_30277.png 1_30_30286.png 1_30_30561.png 1_30_30736.png 1_30_30757.png 1_30_30927.png 1_31_31439.png 1_31_31447.png 1_31_31682.png 1_31_31720.png 1_31_31764.png 1_31_31773.png 1_31_31980.png 1_32_32148.png 1_32_32165.png 1_32_32185.png 1_32_32226.png 1_32_32331.png 1_33_33347.png 1_33_33527.png 1_33_33554.png 1_34_34086.png 1_34_34236.png 1_34_34261.png 1_34_34467.png 1_34_34578.png 1_35_35516.png 1_36_36235.png 1_36_36321.png 1_36_36429.png 1_36_36543.png 1_37_37446.png 1_37_37702.png 1_37_37703.png 1_38_38130.png 1_38_38131.png 1_38_38229.png 1_39_39185.png 1_39_39354.png 1_39_39603.png 1_39_39738.png 1_39_39739.png 1_39_39748.png 1_39_39917.png 1_40_40642.png 1_41_41001.png 1_41_41022.png 1_41_41023.png 1_41_41686.png 1_42_42334.png 1_42_42470.png 1_42_42484.png 1_42_42881.png 1_43_43183.png 1_43_43215.png 1_43_43412.png 1_43_43637.png 1_43_43894.png 1_43_43899.png 1_44_44044.png 1_44_44112.png 1_44_44321.png 1_44_44344.png 1_44_44379.png 1_44_44553.png 1_45_45287.png 1_45_45328.png 1_45_45358.png 1_45_45405.png 1_45_45577.png 1_45_45691.png 1_45_45777.png 1_46_46151.png 1_46_46312.png 1_46_46662.png 1_46_46672.png 1_46_46825.png 1_46_46857.png 1_46_46992.png 1_47_47211.png 1_47_47334.png 1_47_47964.png 1_48_48060.png 1_48_48251.png 1_48_48311.png 1_48_48383.png 1_48_48386.png 1_48_48388.png 1_48_48668.png 1_48_48755.png 1_48_48840.png 1_48_48924.png 1_49_49039.png 1_49_49123.png 1_49_49419.png 1_49_49481.png 1_49_49500.png 1_49_49503.png 1_49_49674.png 1_49_49900.png 1_50_50413.png 1_50_50833.png 1_50_50836.png 1_50_50878.png 1_51_51116.png 1_51_51134.png 1_51_51286.png 1_51_51366.png 1_51_51642.png 1_51_51689.png 1_51_51725.png 1_51_51784.png 1_51_51872.png 1_51_51875.png 1_51_51972.png 1_52_52052.png 1_52_52082.png 1_52_52465.png 1_52_52759.png 1_52_52878.png 1_52_52964.png 1_52_52999.png 1_53_53009.png 1_53_53174.png 1_53_53497.png 1_53_53725.png 1_53_53766.png 1_53_53905.png 1_53_53989.png 1_54_54381.png 1_54_54383.png 1_54_54722.png 1_55_55207.png 1_55_55230.png 1_55_55344.png 1_55_55346.png 1_55_55407.png 1_55_55491.png 1_55_55589.png 1_55_55601.png 1_55_55613.png 1_55_55677.png 1_55_55731.png 1_55_55874.png 1_55_55923.png 1_55_55970.png 1_56_56041.png 1_56_56043.png 1_56_56048.png 1_56_56055.png 1_56_56150.png 1_56_56155.png 1_56_56159.png 1_56_56174.png 1_56_56543.png 1_57_57054.png 1_57_57302.png 1_57_57441.png 1_57_57642.png 1_57_57798.png 1_57_57840.png 1_57_57910.png 1_58_58696.png 1_58_58843.png 1_58_58854.png 1_58_58924.png 1_58_58944.png 1_59_59047.png 1_59_59085.png 1_59_59137.png 1_59_59253.png 1_59_59421.png 1_59_59468.png 1_59_59495.png 1_59_59550.png 1_59_59563.png 1_59_59823.png 1_59_59909.png 1_60_60097.png 1_60_60281.png 1_60_60387.png 1_60_60730.png 1_60_60758.png 1_60_60884.png 1_60_60933.png 1_61_61011.png 1_61_61018.png 1_61_61098.png 1_61_61141.png 1_61_61192.png 1_61_61260.png 1_61_61299.png 1_61_61562.png 1_61_61603.png 1_61_61640.png 1_61_61668.png 1_61_61767.png 1_62_62065.png 1_62_62798.png 1_62_62861.png 1_63_63150.png 1_63_63398.png 1_63_63411.png 1_63_63499.png 1_64_64053.png 1_64_64272.png 1_64_64480.png 1_64_64709.png 1_64_64815.png 1_64_64891.png 1_65_65200.png 1_65_65292.png 1_65_65437.png 1_65_65742.png 1_65_65743.png 1_65_65859.png 1_65_65864.png 1_65_65994.png 1_66_66017.png 1_66_66111.png 1_66_66201.png 1_66_66254.png 1_66_66552.png 1_66_66640.png 1_66_66667.png 1_66_66696.png 1_66_66821.png 1_66_66940.png 1_66_66941.png 1_67_67165.png 2_68_68343.png 2_68_68592.png 2_68_68932.png 2_69_69141.png 2_69_69152.png 2_69_69490.png 2_69_69666.png 2_69_69802.png 2_69_69903.png 2_70_70132.png 2_70_70211.png 2_70_70242.png 2_70_70269.png 2_70_70320.png 2_70_70587.png 2_70_70604.png 2_70_70669.png 2_70_70693.png 2_70_70761.png 2_70_70783.png 2_70_70863.png 2_70_70915.png 2_70_70943.png 2_71_71085.png 2_71_71632.png 2_71_71771.png 2_72_72152.png 2_72_72319.png 2_72_72352.png 2_72_72377.png 2_72_72778.png 2_72_72940.png 2_73_73034.png 2_73_73225.png 2_73_73260.png 2_73_73453.png 2_73_73493.png 2_73_73494.png 2_74_74096.png 2_74_74124.png 2_74_74172.png 2_74_74275.png 2_74_74293.png 2_74_74385.png 2_74_74448.png 2_74_74478.png 2_74_74641.png 2_74_74722.png 2_74_74783.png 2_74_74876.png 2_74_74879.png 2_74_74979.png 2_75_75154.png 2_75_75231.png 2_75_75252.png 2_75_75449.png 2_75_75805.png 2_75_75920.png 2_75_75971.png 2_76_76026.png 2_76_76123.png 2_76_76170.png 2_76_76184.png 2_76_76186.png 2_76_76272.png 2_76_76326.png 2_76_76343.png 2_76_76407.png 2_76_76750.png 2_76_76795.png 2_76_76819.png 2_76_76831.png 2_76_76874.png 2_76_76880.png 2_76_76897.png 2_76_76957.png 2_77_77008.png 2_77_77023.png 2_77_77055.png 2_77_77058.png 2_77_77109.png 2_77_77133.png 2_77_77167.png 2_77_77529.png 2_77_77552.png 2_77_77563.png 2_78_78096.png 2_78_78222.png 2_78_78261.png 2_78_78523.png 2_78_78545.png 2_78_78560.png 2_78_78635.png 2_78_78656.png 2_78_78706.png 2_78_78716.png 2_78_78986.png 2_78_78993.png 2_79_79255.png 2_79_79293.png 2_79_79327.png 2_79_79424.png 2_79_79609.png 2_79_79621.png 2_79_79705.png 2_79_79889.png 2_79_79959.png 2_80_80028.png 2_80_80321.png 2_80_80333.png 2_80_80398.png 2_80_80528.png 2_80_80552.png 2_80_80567.png 2_80_80715.png 2_80_80887.png 2_80_80949.png 2_81_81304.png 2_81_81531.png 2_81_81637.png 2_81_81639.png 2_81_81718.png 2_81_81784.png 2_81_81827.png 2_81_81838.png 2_81_81862.png 2_81_81988.png 2_82_82130.png 2_82_82198.png 2_82_82204.png 2_82_82220.png 2_82_82287.png 2_82_82301.png 2_82_82346.png 2_82_82528.png 2_82_82695.png 2_82_82757.png 2_82_82947.png 2_83_83133.png 2_83_83230.png 2_83_83356.png 2_83_83713.png 2_84_84012.png 2_84_84240.png 2_84_84342.png 2_84_84343.png 2_84_84466.png 2_84_84496.png 2_84_84574.png 2_84_84979.png 2_85_85111.png 2_85_85210.png 2_85_85288.png 2_85_85496.png 2_85_85551.png 2_85_85617.png 2_85_85861.png 2_86_86020.png 2_86_86173.png 2_86_86300.png 2_86_86324.png 2_86_86600.png 2_86_86682.png 2_86_86808.png 2_86_86828.png 2_86_86999.png 2_87_87077.png 2_87_87900.png 2_87_87983.png 2_88_88159.png 2_88_88163.png 2_88_88198.png 2_88_88416.png 2_88_88528.png 2_88_88645.png 2_88_88911.png 2_89_89063.png 2_89_89169.png 2_89_89218.png 2_89_89250.png 2_89_89283.png 2_89_89348.png 2_89_89433.png 2_89_89460.png 2_89_89660.png 2_89_89673.png 2_89_89688.png 2_89_89851.png 2_89_89905.png 2_89_89955.png 2_89_89991.png 2_89_89997.png 2_90_90039.png 2_90_90119.png 2_90_90287.png 2_90_90306.png 2_90_90342.png 2_90_90392.png 2_90_90393.png 2_90_90407.png 2_90_90434.png 2_90_90503.png 2_90_90567.png 2_90_90626.png 2_90_90658.png 2_90_90771.png 2_90_90887.png 2_91_91391.png 2_91_91646.png 2_91_91717.png 2_91_91732.png 2_91_91853.png 2_91_91953.png 2_92_92143.png 2_92_92179.png 2_92_92216.png 2_92_92253.png 2_92_92287.png 2_92_92517.png 2_92_92558.png 2_92_92577.png 2_92_92678.png 2_92_92684.png 2_92_92797.png 2_92_92836.png 2_92_92852.png 2_92_92860.png 2_92_92935.png 2_93_93105.png 2_93_93337.png 2_93_93365.png 2_93_93375.png 2_93_93506.png 2_93_93515.png 2_93_93571.png 2_93_93663.png 2_93_93781.png 2_93_93798.png 2_93_93890.png 2_93_93933.png 2_93_93978.png 2_94_94015.png 2_94_94088.png 3_100_100061.png 3_100_100070.png 3_100_100391.png 3_100_100461.png 3_100_100486.png 3_100_100633.png 3_101_101314.png 3_101_101372.png 3_101_101389.png 3_101_101403.png 3_101_101434.png 3_101_101449.png 3_101_101466.png 3_101_101469.png 3_101_101470.png 3_101_101513.png 3_101_101693.png 3_101_101902.png 3_102_102105.png 3_102_102440.png 3_102_102445.png 3_102_102510.png 3_102_102751.png 3_102_102768.png 3_102_102948.png 3_102_102983.png 3_103_103161.png 3_103_103198.png 3_103_103279.png 3_103_103291.png 3_103_103299.png 3_103_103319.png 3_103_103561.png 3_103_103660.png 3_103_103713.png 3_103_103795.png 3_103_103946.png 3_103_103951.png 3_103_103970.png 3_104_104066.png 3_104_104094.png 3_104_104351.png 3_104_104428.png 3_104_104507.png 3_104_104639.png 3_104_104650.png 3_104_104800.png 3_104_104845.png 3_104_104849.png 3_104_104873.png 3_104_104957.png 3_105_105166.png 3_105_105268.png 3_105_105297.png 3_105_105346.png 3_105_105395.png 3_105_105413.png 3_105_105455.png 3_105_105695.png 3_105_105778.png 3_105_105937.png 3_106_106116.png 3_106_106139.png 3_106_106215.png 3_106_106219.png 3_106_106258.png 3_106_106263.png 3_106_106276.png 3_106_106307.png 3_106_106354.png 3_106_106394.png 3_106_106740.png 3_106_106954.png 3_107_107056.png 3_107_107210.png 3_107_107466.png 3_107_107505.png 3_107_107615.png 3_107_107623.png 3_107_107694.png 3_107_107745.png 3_107_107892.png 3_107_107934.png 3_107_107985.png 3_108_108132.png 3_108_108227.png 3_108_108450.png 3_108_108487.png 3_108_108559.png 3_108_108588.png 3_108_108793.png 3_108_108853.png 3_109_109104.png 3_109_109233.png 3_109_109241.png 3_109_109297.png 3_109_109356.png 3_109_109377.png 3_109_109424.png 3_109_109435.png 3_109_109647.png 3_109_109851.png 3_109_109932.png 3_110_110340.png 3_110_110554.png 3_110_110864.png 3_111_111124.png 3_111_111223.png 3_111_111246.png 3_111_111461.png 3_111_111687.png 3_111_111694.png 3_111_111877.png 3_111_111955.png 3_111_111956.png 3_111_111965.png 3_111_111980.png 3_111_111984.png 3_112_112102.png 3_112_112121.png 3_112_112122.png 3_112_112135.png 3_112_112198.png 3_112_112621.png 3_112_112691.png 3_112_112751.png 3_112_112855.png 3_112_112981.png 3_113_113361.png 3_113_113865.png 3_113_113879.png 3_114_114100.png 3_114_114110.png 3_114_114436.png 3_114_114993.png 3_115_115077.png 3_115_115125.png 3_115_115151.png 3_115_115396.png 3_115_115537.png 3_115_115554.png 3_115_115573.png 3_115_115809.png 3_115_115958.png 3_116_116237.png 3_116_116365.png 3_116_116448.png 3_116_116499.png 3_116_116747.png 3_117_117086.png 3_117_117104.png 3_117_117165.png 3_117_117200.png 3_117_117328.png 3_117_117382.png 3_117_117435.png 3_117_117491.png 3_117_117819.png 3_118_118210.png 3_118_118272.png 3_118_118274.png 3_118_118283.png 3_118_118374.png 3_118_118428.png 3_118_118582.png 3_118_118639.png 3_118_118791.png 3_118_118966.png 3_119_119029.png 3_119_119076.png 3_119_119106.png 3_119_119338.png 3_119_119650.png 3_119_119732.png 3_119_119915.png 3_120_120015.png 3_120_120197.png 3_120_120394.png 3_120_120451.png 3_120_120553.png 3_120_120641.png 3_120_120781.png 3_121_121090.png 3_121_121194.png 3_121_121200.png 3_121_121834.png 3_121_121982.png 3_122_122004.png 3_122_122407.png 3_122_122904.png 3_123_123059.png 3_123_123374.png 3_123_123424.png 3_123_123510.png 3_123_123538.png 3_123_123573.png 3_94_94137.png 3_94_94263.png 3_94_94291.png 3_94_94359.png 3_94_94371.png 3_94_94490.png 3_94_94650.png 3_94_94781.png 3_94_94826.png 3_94_94844.png 3_94_94894.png 3_94_94902.png 3_94_94953.png 3_94_94985.png 3_95_95044.png 3_95_95049.png 3_95_95116.png 3_95_95325.png 3_95_95340.png 3_95_95343.png 3_95_95440.png 3_95_95581.png 3_95_95597.png 3_95_95642.png 3_95_95660.png 3_95_95680.png 3_95_95689.png 3_96_96013.png 3_96_96207.png 3_96_96340.png 3_96_96563.png 3_96_96597.png 3_96_96769.png 3_96_96778.png 3_96_96962.png 3_97_97137.png 3_97_97280.png 3_97_97676.png 3_97_97769.png 3_97_97863.png 3_97_97915.png 3_97_97976.png 3_97_97994.png 3_98_98012.png 3_98_98020.png 3_98_98087.png 3_98_98252.png 3_98_98257.png 3_98_98431.png 3_98_98589.png 3_98_98996.png 3_99_99079.png 3_99_99198.png 3_99_99257.png 3_99_99263.png 3_99_99279.png 3_99_99456.png 3_99_99483.png 3_99_99610.png 3_99_99802.png 4_123_123574.png 4_123_123581.png 4_123_123636.png 4_123_123716.png 4_123_123812.png 4_123_123856.png 4_123_123861.png 4_123_123878.png 4_123_123913.png 4_123_123977.png 4_124_124375.png 4_124_124377.png 4_124_124483.png 4_124_124643.png 4_124_124876.png 4_124_124880.png 4_125_125138.png 4_125_125437.png 4_125_125471.png 4_125_125518.png 4_125_125649.png 4_125_125929.png 4_125_125956.png 4_125_125980.png 4_126_126005.png 4_126_126023.png 4_126_126053.png 4_126_126054.png 4_126_126155.png 4_126_126638.png 4_126_126746.png 4_126_126766.png 4_126_126785.png 4_126_126800.png 4_126_126969.png 4_127_127026.png 4_127_127358.png 4_127_127435.png 4_127_127517.png 4_127_127625.png 4_127_127671.png 4_127_127716.png 4_127_127749.png 4_127_127788.png 4_127_127894.png 4_127_127915.png 4_128_128083.png 4_128_128152.png 4_128_128173.png 4_128_128245.png 4_128_128295.png 4_128_128372.png 4_128_128373.png 4_128_128411.png 4_128_128425.png 4_128_128521.png 4_128_128631.png 4_128_128633.png 4_128_128655.png 4_128_128762.png 4_128_128828.png 4_128_128849.png 4_129_129015.png 4_129_129095.png 4_129_129141.png 4_129_129409.png 4_129_129425.png 4_129_129551.png 4_129_129588.png 4_129_129868.png 4_129_129938.png 4_130_130020.png 4_130_130042.png 4_130_130456.png 4_130_130602.png 4_130_130889.png 4_130_130925.png 4_131_131042.png 4_131_131044.png 4_131_131164.png 4_131_131261.png 4_131_131262.png 4_131_131328.png 4_131_131454.png 4_131_131588.png 4_131_131610.png 4_131_131654.png 4_132_132370.png 4_132_132381.png 4_132_132400.png 4_132_132457.png 4_132_132481.png 4_132_132501.png 4_132_132513.png 4_132_132515.png 4_132_132555.png 4_132_132581.png 4_132_132583.png 4_132_132584.png 4_132_132615.png 4_132_132826.png 4_133_133086.png 4_133_133166.png 4_133_133167.png 4_133_133244.png 4_133_133309.png 4_133_133490.png 4_133_133491.png 4_133_133603.png 4_133_133636.png 4_133_133693.png 4_133_133698.png 4_133_133761.png 4_133_133770.png 4_133_133988.png 4_134_134206.png 4_134_134216.png 4_134_134338.png 4_134_134378.png 4_134_134592.png 4_134_134647.png 4_134_134652.png 4_134_134689.png 4_134_134720.png 4_134_134797.png 4_134_134940.png 4_134_134943.png 4_134_134970.png 4_135_135051.png 4_135_135123.png 4_135_135184.png 4_135_135296.png 4_135_135739.png 4_135_135781.png 4_135_135795.png 4_136_136057.png 4_136_136256.png 4_136_136419.png 4_136_136517.png 4_136_136550.png 4_136_136593.png 4_136_136640.png 4_136_136642.png 4_136_136917.png 4_137_137110.png 4_137_137215.png 4_137_137256.png 4_137_137270.png 4_137_137457.png 4_137_137663.png 4_137_137833.png 4_137_137858.png 4_138_138100.png 4_138_138232.png 4_138_138371.png 4_138_138720.png 4_138_138730.png 4_138_138864.png 4_138_138880.png 4_138_138914.png 4_138_138932.png 4_138_138935.png 4_138_138939.png 4_139_139056.png 4_139_139062.png 4_139_139102.png 4_139_139123.png 4_139_139232.png 4_139_139240.png 4_139_139244.png 4_139_139249.png 4_139_139325.png 4_139_139369.png 4_139_139482.png 4_139_139513.png 4_139_139556.png 4_139_139680.png 4_140_140067.png 4_140_140136.png 4_140_140140.png 4_140_140266.png 4_140_140285.png 4_140_140318.png 4_140_140347.png 4_140_140696.png 4_140_140699.png 4_140_140735.png 4_140_140766.png 4_141_141132.png 4_141_141284.png 4_141_141406.png 4_141_141471.png 4_141_141474.png 4_141_141480.png 4_141_141483.png 4_141_141718.png 4_141_141727.png 4_141_141743.png 4_141_141769.png 4_141_141771.png 4_141_141791.png 4_141_141883.png 4_141_141894.png 4_141_141905.png 4_141_141907.png 4_141_141914.png 4_141_141919.png 4_142_142037.png 4_142_142102.png 4_142_142115.png 4_142_142233.png 4_142_142237.png 4_142_142258.png 4_142_142301.png 4_142_142364.png 4_142_142469.png 4_142_142509.png 4_142_142520.png 4_142_142527.png 4_142_142537.png 4_142_142541.png 4_142_142588.png 4_142_142598.png 4_142_142627.png 4_142_142628.png 4_142_142635.png 4_142_142642.png 4_142_142644.png 4_142_142658.png 4_142_142677.png 4_142_142715.png 4_142_142765.png 4_142_142848.png 4_142_142887.png 4_142_142894.png 4_142_142916.png 4_143_143056.png 4_143_143078.png 4_143_143198.png 4_143_143253.png 4_143_143505.png 4_143_143609.png 4_143_143630.png 4_143_143691.png 4_143_143735.png 4_143_143747.png 4_143_143757.png 4_143_143828.png 4_143_143913.png 4_143_143962.png 4_144_144178.png 4_144_144374.png 4_144_144408.png 4_144_144439.png 4_144_144500.png 4_144_144682.png 4_144_144785.png 4_144_144961.png 5_145_145275.png 5_145_145322.png 5_145_145486.png 5_145_145523.png 5_145_145717.png 5_145_145839.png 5_146_146158.png 5_146_146319.png 5_146_146332.png 5_146_146464.png 5_146_146546.png 5_146_146551.png 5_146_146599.png 5_146_146660.png 5_146_146838.png 5_147_147030.png 5_147_147132.png 5_147_147338.png 5_147_147342.png 5_147_147394.png 5_147_147693.png 5_147_147694.png 5_147_147712.png 5_147_147898.png 5_148_148059.png 5_148_148067.png 5_148_148271.png 5_148_148420.png 5_148_148438.png 5_148_148593.png 5_148_148662.png 5_148_148710.png 5_148_148954.png 5_149_149055.png 5_149_149153.png 5_149_149828.png 5_149_149842.png 5_149_149858.png 5_150_150104.png 5_150_150186.png 5_150_150259.png 5_150_150306.png 5_150_150486.png 5_150_150513.png 5_150_150624.png 5_150_150633.png 5_150_150723.png 5_150_150979.png 5_151_151134.png 5_151_151223.png 5_151_151244.png 5_151_151463.png 5_151_151526.png 5_151_151548.png 5_151_151641.png 5_151_151771.png 5_151_151933.png 5_152_152103.png 5_152_152125.png 5_152_152191.png 5_152_152335.png 5_152_152353.png 5_152_152356.png 5_152_152424.png 5_152_152463.png 5_152_152659.png 5_152_152697.png 5_152_152703.png 5_152_152955.png 5_153_153003.png 5_153_153554.png 5_153_153790.png 5_153_153805.png 5_153_153947.png 5_153_153969.png 5_154_154001.png 5_154_154148.png 5_154_154292.png 5_154_154297.png 5_154_154324.png 5_154_154588.png 5_154_154914.png 5_155_155026.png 5_155_155028.png 5_155_155096.png 5_155_155113.png 5_155_155145.png 5_155_155196.png 5_155_155204.png 5_155_155206.png 5_155_155260.png 5_155_155261.png 5_155_155451.png 5_155_155658.png 5_155_155745.png 5_155_155983.png 5_156_156109.png 5_156_156160.png 5_156_156251.png 5_156_156260.png 5_156_156392.png 5_156_156422.png 5_156_156556.png 5_156_156576.png 5_156_156589.png 5_157_157199.png 5_157_157291.png 5_157_157343.png 5_157_157357.png 5_157_157409.png 5_157_157466.png 5_157_157483.png 5_157_157619.png 5_157_157620.png 5_158_158013.png 5_158_158155.png 5_158_158233.png 5_158_158331.png 5_158_158836.png 5_159_159091.png 5_159_159161.png 5_159_159505.png 5_159_159663.png 5_159_159913.png 5_159_159927.png 5_159_159951.png 5_160_160012.png 5_160_160069.png 5_160_160098.png 5_160_160128.png 5_160_160230.png 5_160_160233.png 5_160_160300.png 5_160_160330.png 5_160_160350.png 5_160_160390.png 5_160_160391.png 5_160_160395.png 5_160_160397.png 5_160_160431.png 5_160_160446.png 5_160_160469.png 5_160_160595.png 5_160_160680.png 5_160_160822.png 5_160_160866.png 5_160_160877.png 5_160_160955.png 5_161_161112.png 5_161_161126.png 5_161_161211.png 5_161_161406.png 5_161_161761.png 5_161_161904.png 5_161_161927.png 5_161_161983.png 5_161_161998.png 5_162_162074.png 5_162_162119.png 5_162_162125.png 5_162_162166.png 5_162_162198.png 5_162_162225.png 5_162_162354.png 5_162_162366.png 5_162_162402.png 5_162_162408.png 5_162_162434.png 5_162_162516.png 5_162_162517.png 5_162_162556.png 5_162_162776.png 5_162_162862.png 5_162_162967.png 5_163_163052.png 5_163_163129.png 5_163_163211.png 5_163_163223.png 5_163_163256.png 5_163_163434.png 5_163_163456.png 5_163_163466.png 5_163_163594.png 5_163_163607.png 5_163_163674.png 5_163_163696.png 5_163_163729.png 5_163_163966.png 5_163_163989.png 5_164_164002.png 5_164_164225.png 5_164_164379.png 5_164_164390.png 5_164_164391.png 5_164_164407.png 5_164_164442.png 5_164_164467.png 5_164_164472.png 5_164_164583.png 5_164_164617.png 5_164_164621.png 5_164_164643.png 5_164_164666.png 5_164_164713.png 5_164_164763.png 5_164_164983.png 5_165_165017.png 5_165_165210.png 5_165_165607.png 5_165_165988.png 5_166_166065.png 5_166_166071.png 5_166_166189.png 5_166_166319.png 5_166_166408.png 5_166_166575.png 5_166_166681.png 5_166_166694.png 5_166_166949.png 5_167_167128.png 5_167_167168.png 5_167_167302.png 5_167_167321.png 5_167_167613.png 5_167_167742.png 5_167_167878.png 5_167_167879.png 5_167_167975.png 5_167_167984.png 5_167_167989.png 5_167_167993.png 5_168_168126.png 5_168_168177.png 5_168_168234.png 5_168_168242.png 5_168_168383.png 5_168_168407.png 5_168_168671.png 5_168_168673.png 5_168_168878.png 5_168_168888.png 5_168_168995.png 5_169_169035.png 5_169_169042.png 5_169_169043.png 6_169_169128.png 6_169_169288.png 6_169_169361.png 6_169_169367.png 6_169_169368.png 6_169_169397.png 6_169_169491.png 6_169_169526.png 6_169_169936.png 6_170_170047.png 6_170_170061.png 6_170_170079.png 6_170_170155.png 6_170_170160.png 6_170_170170.png 6_170_170273.png 6_170_170471.png 6_170_170722.png 6_170_170745.png 6_170_170862.png 6_170_170880.png 6_170_170933.png 6_171_171024.png 6_171_171096.png 6_171_171163.png 6_171_171274.png 6_171_171354.png 6_171_171556.png 6_171_171590.png 6_171_171631.png 6_171_171866.png 6_171_171885.png 6_171_171936.png 6_172_172511.png 6_172_172514.png 6_172_172668.png 6_172_172688.png 6_172_172693.png 6_172_172700.png 6_172_172802.png 6_172_172922.png 6_173_173010.png 6_173_173119.png 6_173_173247.png 6_173_173567.png 6_173_173657.png 6_173_173705.png 6_173_173708.png 6_173_173740.png 6_173_173794.png 6_173_173847.png 6_174_174049.png 6_174_174050.png 6_174_174086.png 6_174_174252.png 6_174_174261.png 6_174_174336.png 6_174_174349.png 6_174_174394.png 6_174_174476.png 6_174_174534.png 6_174_174563.png 6_174_174747.png 6_174_174769.png 6_174_174805.png 6_174_174973.png 6_174_174986.png 6_175_175009.png 6_175_175026.png 6_175_175106.png 6_175_175269.png 6_175_175275.png 6_175_175466.png 6_175_175481.png 6_175_175506.png 6_175_175647.png 6_175_175676.png 6_175_175726.png 6_175_175967.png 6_176_176075.png 6_176_176314.png 6_176_176357.png 6_176_176381.png 6_176_176386.png 6_176_176526.png 6_176_176617.png 6_176_176649.png 6_176_176923.png 6_176_176982.png 6_177_177110.png 6_177_177129.png 6_177_177213.png 6_177_177229.png 6_177_177302.png 6_177_177307.png 6_177_177397.png 6_177_177415.png 6_177_177722.png 6_178_178076.png 6_178_178122.png 6_178_178143.png 6_178_178187.png 6_178_178275.png 6_178_178277.png 6_178_178279.png 6_178_178280.png 6_178_178281.png 6_178_178282.png 6_178_178399.png 6_178_178407.png 6_178_178467.png 6_178_178511.png 6_178_178685.png 6_178_178822.png 6_178_178865.png 6_178_178879.png 6_178_178884.png 6_178_178885.png 6_178_178980.png 6_179_179030.png 6_179_179090.png 6_179_179207.png 6_179_179287.png 6_179_179324.png 6_179_179398.png 6_179_179416.png 6_179_179431.png 6_179_179434.png 6_179_179438.png 6_179_179605.png 6_179_179654.png 6_179_179783.png 6_180_180004.png 6_180_180008.png 6_180_180080.png 6_180_180251.png 6_180_180282.png 6_180_180577.png 6_180_180703.png 6_180_180705.png 6_180_180781.png 6_180_180885.png 6_181_181006.png 6_181_181152.png 6_181_181182.png 6_181_181344.png 6_181_181382.png 6_181_181415.png 6_181_181443.png 6_181_181787.png 6_181_181837.png 6_181_181841.png 6_181_181842.png 6_181_181845.png 6_181_181907.png 6_181_181909.png 6_181_181932.png 6_181_181933.png 6_181_181934.png 6_181_181938.png 6_181_181950.png 6_181_181951.png 6_181_181954.png 6_182_182027.png 6_182_182070.png 6_182_182191.png 6_182_182214.png 6_182_182221.png 6_182_182424.png 6_182_182480.png 6_182_182498.png 6_182_182531.png 6_182_182532.png 6_182_182566.png 6_182_182722.png 6_182_182890.png 6_183_183024.png 6_183_183051.png 6_183_183192.png 6_183_183414.png 6_183_183464.png 6_183_183467.png 6_183_183539.png 6_183_183614.png 6_183_183629.png 6_183_183730.png 6_183_183779.png 6_183_183795.png 6_183_183803.png 6_183_183804.png 6_183_183829.png 6_183_183831.png 6_183_183879.png 6_184_184112.png 6_184_184204.png 6_184_184223.png 6_184_184238.png 6_184_184310.png 6_184_184312.png 6_184_184313.png 6_184_184326.png 6_184_184334.png 6_184_184356.png 6_184_184364.png 6_184_184366.png 6_184_184376.png 6_184_184487.png 6_184_184513.png 6_184_184775.png 6_184_184979.png 6_185_185045.png 6_185_185046.png 6_185_185092.png 6_185_185101.png 6_185_185118.png 6_185_185147.png 6_185_185156.png 6_185_185166.png 6_185_185184.png 6_185_185261.png 6_185_185267.png 6_185_185320.png 6_185_185436.png 6_185_185479.png 6_185_185581.png 6_185_185592.png 6_185_185714.png 6_185_185883.png 6_185_185884.png 6_185_185951.png 6_186_186022.png 6_186_186041.png 6_186_186109.png 6_186_186111.png 6_186_186120.png 6_186_186216.png 6_186_186294.png 6_186_186299.png 6_186_186340.png 6_186_186617.png 6_186_186637.png 6_186_186717.png 6_186_186738.png 6_186_186799.png 6_186_186911.png 6_186_186937.png 6_186_186978.png 6_187_187026.png 6_187_187039.png 6_187_187060.png 6_187_187068.png 6_187_187092.png 6_187_187187.png 6_187_187192.png 6_187_187270.png 6_187_187587.png 7_187_187644.png 7_187_187772.png 7_187_187858.png 7_187_187877.png 7_187_187891.png 7_187_187965.png 7_187_187975.png 7_188_188009.png 7_188_188016.png 7_188_188041.png 7_188_188053.png 7_188_188076.png 7_188_188150.png 7_188_188152.png 7_188_188221.png 7_188_188335.png 7_188_188375.png 7_188_188443.png 7_188_188612.png 7_188_188629.png 7_188_188671.png 7_188_188821.png 7_188_188858.png 7_188_188878.png 7_188_188900.png 7_188_188996.png 7_189_189052.png 7_189_189055.png 7_189_189111.png 7_189_189257.png 7_189_189291.png 7_189_189358.png 7_189_189395.png 7_189_189447.png 7_189_189451.png 7_189_189652.png 7_189_189653.png 7_189_189656.png 7_189_189668.png 7_189_189749.png 7_190_190106.png 7_190_190268.png 7_190_190361.png 7_190_190595.png 7_190_190605.png 7_190_190623.png 7_190_190778.png 7_190_190783.png 7_190_190902.png 7_190_190969.png 7_190_190999.png 7_191_191008.png 7_191_191009.png 7_191_191154.png 7_191_191174.png 7_191_191272.png 7_191_191341.png 7_191_191373.png 7_191_191513.png 7_192_192101.png 7_192_192173.png 7_192_192350.png 7_192_192365.png 7_192_192428.png 7_192_192719.png 7_192_192856.png 7_192_192869.png 7_192_192893.png 7_192_192928.png 7_193_193020.png 7_193_193104.png 7_193_193139.png 7_193_193991.png 7_194_194027.png 7_194_194183.png 7_194_194203.png 7_194_194215.png 7_194_194229.png 7_194_194243.png 7_194_194269.png 7_194_194797.png 7_194_194833.png 7_194_194899.png 7_194_194993.png 7_195_195176.png 7_195_195198.png 7_195_195416.png 7_195_195611.png 7_195_195718.png 7_195_195812.png 7_195_195882.png 7_196_196096.png 7_196_196240.png 7_196_196386.png 7_196_196614.png 7_196_196631.png 7_196_196775.png 7_196_196776.png 7_196_196820.png 7_196_196980.png 7_197_197020.png 7_197_197087.png 7_197_197102.png 7_197_197113.png 7_197_197186.png 7_197_197349.png 7_197_197390.png 7_197_197411.png 7_197_197514.png 7_197_197520.png 7_197_197589.png 7_197_197682.png 7_197_197687.png 7_197_197876.png 7_197_197996.png 7_198_198044.png 7_198_198073.png 7_198_198095.png 7_198_198179.png 7_198_198183.png 7_198_198394.png 7_198_198400.png 7_198_198494.png 7_198_198638.png 7_198_198640.png 7_198_198728.png 7_198_198748.png 7_198_198918.png 7_198_198953.png 7_198_198958.png 7_198_198998.png 7_199_199045.png 7_199_199059.png 7_199_199241.png 7_199_199273.png 7_199_199281.png 7_199_199309.png 7_199_199366.png 7_199_199442.png 7_199_199642.png 7_199_199643.png 7_199_199669.png 7_199_199691.png 7_199_199692.png 7_199_199936.png 7_199_199952.png 7_200_200012.png 7_200_200378.png 7_200_200384.png 7_200_200443.png 7_200_200555.png 7_200_200745.png 7_200_200941.png 7_201_201035.png 7_201_201126.png 7_201_201213.png 7_201_201292.png 7_201_201326.png 7_201_201366.png 7_201_201394.png 7_201_201412.png 7_201_201516.png 7_201_201545.png 7_201_201725.png 7_201_201743.png 7_201_201758.png 7_202_202033.png 7_202_202075.png 7_202_202157.png 7_202_202166.png 7_202_202183.png 7_202_202235.png 7_202_202280.png 7_202_202289.png 7_202_202293.png 7_202_202308.png 7_202_202345.png 7_202_202422.png 7_202_202435.png 7_202_202443.png 7_202_202489.png 7_202_202553.png 7_202_202565.png 7_202_202576.png 7_202_202587.png 7_202_202613.png 7_202_202637.png 7_202_202708.png 7_202_202723.png 7_202_202920.png 7_203_203069.png 7_203_203099.png 7_203_203252.png 7_203_203453.png 7_203_203526.png 7_203_203760.png 7_203_203849.png 7_203_203851.png 7_203_203897.png 7_204_204017.png 7_204_204115.png 7_204_204201.png 7_204_204261.png 7_204_204300.png 7_204_204588.png 7_204_204599.png 7_204_204609.png 7_204_204661.png 7_204_204745.png 7_204_204802.png 7_204_204804.png 7_204_204947.png 7_205_205028.png 7_205_205040.png 7_205_205157.png 7_205_205162.png 7_205_205174.png 7_205_205250.png 7_205_205328.png 7_205_205332.png 7_205_205363.png 7_205_205373.png 7_205_205377.png 7_205_205420.png 7_205_205479.png 7_205_205500.png 7_205_205544.png 7_205_205643.png 7_205_205702.png 7_205_205703.png 7_205_205721.png 7_205_205900.png 7_206_206022.png 7_206_206026.png 7_206_206284.png 7_206_206330.png 7_206_206481.png 7_206_206548.png 7_206_206606.png 7_206_206631.png 7_206_206703.png 9_100075.png 9_100080.png 9_100098.png 9_100633.png 9_100999.png 9_102_0260.png 9_104022.png 9_105053.png 9_108028.png 9_108070.png 9_108076.png 9_108_0873.png 9_108_0881.png 9_108_0883.png 9_108_0892.png 9_109034.png 9_109053.png 9_109_0936.png 9_110_1002.png 9_110_1027.png 9_110_1028.png 9_110_1030.png 9_110_1072.png 9_111_1136.png 9_111_1139.png 9_111_1153.png 9_111_1162.png 9_112268.png 9_112449.png 9_112515.png 9_112_1276.png 9_113016.png 9_113044.png 9_113418.png 9_114_1428.png 9_114_1429.png 9_114_1431.png 9_114_1435.png 9_114_1439.png 9_114_1443.png 9_114_1446.png 9_114_1447.png 9_114_1464.png 9_114_1467.png 9_115479.png 9_115_1580.png 9_115_1581.png 9_116758.png 9_117003.png 9_117200.png 9_117_1751.png 9_117_1752.png 9_117_1753.png 9_117_1754.png 9_117_1757.png 9_118035.png 9_118_1869.png 9_118_1875.png 9_118_1878.png 9_118_1883.png 9_118_1900.png 9_119_1936.png 9_12084.png 9_122048.png 9_124_2500.png 9_125_2501.png 9_126007.png 9_126_2677.png 9_127_2790.png 9_131260.png 9_134008.png 9_138078.png 9_140061.png 9_140075.png 9_140827.png 9_145014.png 9_145053.png 9_15001.png 9_15002.png 9_15036.png 9_15041.png 9_15063.png 9_15074.png 9_15087.png 9_15088.png 9_15120.png 9_15122.png 9_15151.png 9_15200.png 9_15213.png 9_152_5246.png 9_153_5325.png 9_153_5328.png 9_153_5366.png 9_153_5398.png 9_155060.png 9_156079.png 9_157_5727.png 9_158_5839.png 9_159029.png 9_16052.png 9_160_6083.png 9_161062.png 9_161_6143.png 9_161_6145.png 9_161_6146.png 9_162_6230.png 9_163014.png 9_163062.png 9_164_6484.png 9_166081.png 9_166_6650.png 9_166_6654.png 9_170054.png 9_17050.png 9_17077.png 9_17118.png 9_17188.png 9_172032.png 9_17212.png 9_17238.png 9_17249.png 9_17262.png 9_17282.png 9_17332.png 9_17336.png 9_175_7551.png 9_175_7583.png 9_176019.png 9_178_7847.png 9_178_7848.png 9_179_7945.png 9_179_7946.png 9_180_8015.png 9_180_8024.png 9_180_8028.png 9_180_8044.png 9_183_8309.png 9_183_8314.png 9_183_8318.png 9_183_8320.png 9_183_8321.png 9_183_8322.png 9_183_8336.png 9_183_8395.png 9_184_8494.png 9_187083.png 9_188063.png 9_188091.png 9_189011.png 9_19021.png 9_196015.png 9_198054.png 9_202012.png 9_209070.png 9_21077.png 9_22001.png 9_22013.png 9_22029.png 9_22104.png 9_227040.png 9_23001.png 9_23002.png 9_23025.png 9_23051.png 9_23074.png 9_23080.png 9_23104.png 9_23117.png 9_239007.png 9_239096.png 9_24004.png 9_246016.png 9_260081.png 9_26926127.forsythia1.png 9_26995738.ibiris1.png 9_271008.png 9_28010.png 9_28026.png 9_28052.png 9_28080.png 9_28117.png 9_28162.png 9_28180.png 9_28195.png 9_28198.png 9_28210.png 9_28244.png 9_28278.png 9_28316.png 9_28579914.05.png 9_291000.png 9_299091.png 9_304034.png 9_304074.png 9_306005.png 9_309004.png 9_317080.png 9_326038.png 9_35008.png 9_35010.png 9_35058.png 9_353013.png 9_36253.png 9_37073.png 9_376020.png 9_41025.png 9_41069.png 9_42044.png 9_42049.png 9_43070.png 9_45096.png 9_46076.png 9_50314.png 9_53915809.20051221_27lifeguard.png 9_54375701.bearfishimg_6001640.png 9_58060.png 9_61086.png 9_62096.png 9_65010.png 9_65074.png 9_76053.png 9_78004.png 9_8049.png 9_8143.png 9_85048.png 9_87046.png 9_90076.png 9_95_95660.png 9_97017.png 9_98350.png 9_c_141.png 9_c_146.png 9_image_0004.png 9_image_0014.png 9_image_0015.png 9_image_0025.png 9_image_0027.png 9_image_0036.png 9_image_0039.png 9_image_0044.png 9_image_0054.png 9_image_0057.png 9_image_0058.png 9_image_0068.png 9_image_0072.png 9_image_0083.png 9_image_0086.png 9_image_0103.png 9_pushpin.png 9_ss07002.png 9_ss07033.png 9_ss07036.png 9_ss07080.png 9_ss07085.png 0_0_272.png 0_0_307.png 0_0_355.png 0_0_775.png 0_0_899.png 0_10_10536.png 0_10_10600.png 0_11_11060.png 0_11_11111.png 0_11_11136.png 0_11_11263.png 0_11_11297.png 0_11_11309.png 0_11_11328.png 0_11_11346.png 0_11_11425.png 0_11_11561.png 0_11_11571.png 0_11_11650.png 0_11_11852.png 0_11_11881.png 0_11_11945.png 0_11_11981.png 0_12_12144.png 0_12_12171.png 0_12_12344.png 0_12_12424.png 0_12_12435.png 0_12_12484.png 0_12_12518.png 0_12_12619.png 0_12_12649.png 0_12_12793.png 0_12_12845.png 0_12_12892.png 0_13_13027.png 0_13_13198.png 0_13_13331.png 0_13_13347.png 0_13_13386.png 0_13_13400.png 0_13_13515.png 0_13_13553.png 0_13_13885.png 0_14_14597.png 0_14_14651.png 0_14_14747.png 0_14_14831.png 0_14_14926.png 0_14_14991.png 0_15_15022.png 0_15_15030.png 0_15_15663.png 0_15_15763.png 0_15_15844.png 0_15_15857.png 0_15_15859.png 0_15_15876.png 0_15_15880.png 0_15_15935.png 0_15_15980.png 0_16_16196.png 0_16_16223.png 0_16_16283.png 0_16_16627.png 0_16_16704.png 0_16_16768.png 0_16_16801.png 0_16_16905.png 0_16_16940.png 0_17_17057.png 0_17_17275.png 0_17_17350.png 0_17_17352.png 0_17_17525.png 0_17_17637.png 0_17_17674.png 0_18_18160.png 0_18_18310.png 0_18_18562.png 0_18_18636.png 0_18_18653.png 0_18_18723.png 0_18_18927.png 0_18_18928.png 0_18_18930.png 0_18_18957.png 0_19_19011.png 0_19_19025.png 0_19_19029.png 0_19_19068.png 0_19_19090.png 0_19_19137.png 0_19_19193.png 0_19_19249.png 0_19_19429.png 0_19_19593.png 0_19_19773.png 0_1_1004.png 0_1_1288.png 0_1_1339.png 0_1_1409.png 0_1_1427.png 0_1_1459.png 0_1_1611.png 0_1_1650.png 0_1_1728.png 0_1_1934.png 0_20_20573.png 0_20_20656.png 0_20_20783.png 0_20_20937.png 0_21_21244.png 0_21_21299.png 0_21_21394.png 0_21_21781.png 0_22_22218.png 0_22_22234.png 0_22_22396.png 0_22_22838.png 0_22_22933.png 0_23_23683.png 0_23_23697.png 0_23_23748.png 0_23_23817.png 0_23_23934.png 0_24_24177.png 0_24_24196.png 0_24_24209.png 0_24_24237.png 0_24_24256.png 0_24_24390.png 0_24_24453.png 0_24_24552.png 0_24_24759.png 0_24_24771.png 0_24_24861.png 0_24_24918.png 0_24_24926.png 0_24_24965.png 0_25_25000.png 0_25_25057.png 0_25_25064.png 0_2_2268.png 0_2_2276.png 0_2_2301.png 0_2_2310.png 0_2_2397.png 0_2_2580.png 0_2_2741.png 0_2_2756.png 0_3_3183.png 0_3_3294.png 0_3_3360.png 0_3_3362.png 0_3_3429.png 0_3_3433.png 0_3_3514.png 0_3_3524.png 0_3_3551.png 0_3_3610.png 0_3_3732.png 0_3_3797.png 0_3_3835.png 0_3_3965.png 0_4_4283.png 0_4_4328.png 0_4_4430.png 0_4_4828.png 0_5_5010.png 0_5_5091.png 0_5_5255.png 0_5_5291.png 0_5_5303.png 0_5_5318.png 0_5_5335.png 0_5_5463.png 0_5_5586.png 0_5_5785.png 0_5_5880.png 0_5_5887.png 0_6_6048.png 0_6_6377.png 0_6_6378.png 0_6_6646.png 0_6_6693.png 0_6_6744.png 0_6_6772.png 0_6_6825.png 0_6_6913.png 0_7_7167.png 0_7_7478.png 0_7_7609.png 0_7_7804.png 0_7_7819.png 0_7_7821.png 0_7_7904.png 0_7_7917.png 0_8_8040.png 0_8_8109.png 0_8_8151.png 0_8_8238.png 0_8_8248.png 0_8_8711.png 0_8_8859.png 0_8_8860.png 0_8_8885.png 0_9_9012.png 0_9_9353.png 0_9_9398.png 0_9_9453.png 0_9_9659.png 0_9_9799.png 0_9_9938.png 10_00000013_008.png 10_00000014_010.png 10_00000014_016.png 10_00000027_009.png 10_00000069_018.png 10_00000076_017.png 10_00000083_002.png 10_00000085_008.png 10_00000093_007.png 10_108765872_c1d462aa15.png 10_111078578_967ba44a4e.png 10_115003816_5268a2536a.png 10_135418000_4666dec3c4.png 10_137117726_7d0ef6a2f1.png 10_144439584_97e8823d39.png 10_1557631_c5d89caa41.png 10_156234186_1b16bc540f.png 10_161208056_248b2c2ab6.png 10_169329862_b3b297c7a9.png 10_188491483_a44e3d6e02.png 10_192447210_936a00b749.png 10_197325318_cceafb4122.png 10_200477093_158e8132c2.png 10_201960469_f9110392d6.png 10_206775277_55350e13b2.png 10_215268421_d9e8d9a11d.png 10_230453144_086f387b70.png 10_236916255_307bc2272c.png 10_246111556_0fc2277e24.png 10_252237817_b4d44ff2f7.png 10_257305705_33f7249c08.png 10_260120621_66276c1d6d.png 10_260372546_03d18a4e9e.png 10_261412567_6cfa782a7e.png 10_261964240_8eea8f76d8.png 10_262725611_c3ce2b827e.png 10_263409221_258fe30291.png 10_264482836_a97bd01e84.png 10_264484124_5edf738e12.png 10_264977628_cde9f779bc.png 10_265627532_6350797cbb.png 10_266623585_51b4e13abb.png 10_266774040_f40061481f.png 10_267001913_e22cf985d5.png 10_267412727_1727822888.png 10_267466377_0af5d33261.png 10_267561242_d29d940853.png 10_267661663_55aa54cc71.png 10_267687096_fc062d2623.png 10_267687692_468e936114.png 10_267687863_2b911db6b9.png 10_267688590_6ab2c60659.png 10_267689095_cc345739aa.png 10_267690183_224c509150.png 10_267692977_62b9a260bb.png 10_267693197_b9d8dd822d.png 10_267693828_c8c86c50e0.png 10_267744690_ac99310c04.png 10_267779210_3d38f9a927.png 10_267780676_e2346e581a.png 10_267781192_ae29cb453b.png 10_267781224_2039c2b3fb.png 10_267781270_9648a04e83.png 10_267781788_2765356aeb.png 10_267782578_6f209f0815.png 10_267942018_4013a57095.png 10_267960073_30075b5290.png 10_267975486_b9cd08a46f.png 10_267976432_b4e9429cff.png 10_267976435_62994d7dbe.png 10_267987584_e90efadcf1.png 10_268001846_db119074c8.png 10_268014760_9676395938.png 10_268014977_b47bfb117d.png 10_268026286_c48d28c944.png 10_268033410_d278ddd024.png 10_268077139_f4250e6df3.png 10_268090450_7103d5475d.png 10_268096162_0b812ff56c.png 10_268101001_cb023c75dc.png 10_268113688_6486df5217.png 10_268118418_3053caa1d7.png 10_268122508_c361b4db6c.png 10_268127478_dfb1265d0a.png 10_268129990_04e858de05.png 10_268136909_f2dc3bee43.png 10_268139718_1a9192a52e.png 10_268155424_8f0e352947.png 10_268155490_7dbe72fa01.png 10_268158274_128ba8610b.png 10_268167939_2e6847d767.png 10_268168007_aba64bbe7a.png 10_268168010_d3852a2e6e.png 10_268168390_affe478866.png 10_268169446_590c2cc58e.png 10_268172530_e440f57dcb.png 10_268173785_ddf856ecf5.png 10_268179190_1a6efc3d39.png 10_268180917_4556a9bb75.png 10_268185248_9471d517d0.png 10_268204981_fe60d55f2d.png 10_268207227_f985bd3500.png 10_268211019_270132655c.png 10_268212162_6982292e19.png 10_268214893_d72f1df309.png 10_268217543_52a2013e8c.png 10_268231155_a6210a13b7.png 10_268231902_1586ec90e2.png 10_268239670_5f96290ab3.png 10_268252475_3716432836_m.png 10_268253251_48e9b94730.png 10_268257348_58c33fa2d3.png 10_268262571_a713359657.png 10_268273406_d54a0783ee.png 10_268276931_3412319482.png 10_268278759_66f39a7eb8.png 10_268283673_7b0903063a.png 10_268283809_06f7e47848.png 10_268283974_7a8c597b0e.png 10_268284434_26bb088f22.png 10_268284889_853fac59b5.png 10_268304754_2c3987b247.png 10_268307429_6d09236528.png 10_28664556.moe.png 10_29187800.img_5050.png 10_29421652.pict1700.png 10_29566017.littledave.png 10_29665117_112d6360cd.png 10_29912713.stream5sm.png 10_34312677.foalhead.png 10_36687635.gpzoo_przewalski.png 10_38139938.pbdogs10.png 10_38173560.pbcats02.png 10_38173647.pbcats06.png 10_38527623_0542f07e1c.png 10_39459084.img_0498.png 10_40837349.dsc_0822.png 10_41105485.05_0226_aabbspoth2.png 10_42230402.p1010014_x.png 10_43047581.ds20050506_0127awfcat.png 10_43097193.p5050256.png 10_43642598.pasqueflowerpasque6.png 10_43642599.pasqueflowerpasque7.png 10_43642602.pasqueflowerpasqueflowerearlyeve2.png 10_44859864.1.png 10_44859870.4.png 10_44859876.7.png 10_44859877.10.png 10_44859879.11.png 10_45873784.happychap.png 10_46084071.commontern1.png 10_46642874.roadingjuly09001.png 10_49513253.0509180001.fsn.png 10_49513257.0509180005.fsdp.png 10_49513258.0509180006.fsdm.png 10_49930782.initiationjunior1123.png 10_49930787.initiationjunior1127.png 10_49930789.initiationjunior1130.png 10_49930796.initiationjunior1142.png 10_49930797.initiationjunior1143.png 10_49930798.initiationjunior1144.png 10_49930803.initiationjunior1149.png 10_49930804.initiationjunior1150.png 10_49930807.initiationjunior1153.png 10_52202391.dscn0138.png 10_52202392.dscn0140.png 10_52202419.dscn0189.png 10_52202425.dscn0197.png 10_52202427.dscn0203.png 10_52202430.dscn0206.png 10_52202441.dscn0222.png 10_52202445.dscn0226.png 10_52202447.dscn0228.png 10_52202453.dscn0243.png 10_52202457.dscn0251.png 10_52202465.dscn0265.png 10_52202466.dscn0266.png 10_52202492.dscn0308.png 10_52523077.horse_of_horse_shoes_079.png 10_54628916.jan8_06_536.png 10_54628920.jan8_06_540.png 10_54628932.jan8_06_552.png 10_55651847.carlystreetweballey2.png 10_55768641.checkers22.png 10_57572381.cutesofss.png 10_58229768.20060405022_object_of_fancy.png 10_59242003.img_4193.png 10_59271038.138_3822.png 10_59271039.140_4028.png 10_60866705.butterfly.png 10_61749596.jpvis3ea.redtailedhawkjuvenile.png 10_63193989.iibp3dda.fritillary.png 10_63359627.mq7fmcxg._mg_4055editfromrawsized.png 10_63451013.9howa94v.turkeyvulture.png 10_63827110.laoeb6po.png 10_66830584.jsnb9pmi.sept10_06_704.png 10_66830586.6whtssec.sept10_06_705.png 10_66830612.jwrzvoka.sept10_06_728.png 10_67444.png 10_68285988.3ddkraxw.toughtimeskid.png 10_75671582_152bf239ef.png 10_98048098_af08c0f2df.png 1_25_25191.png 1_25_25246.png 1_25_25406.png 1_25_25451.png 1_25_25580.png 1_25_25895.png 1_25_25953.png 1_26_26013.png 1_26_26230.png 1_26_26236.png 1_26_26256.png 1_26_26335.png 1_26_26532.png 1_26_26643.png 1_26_26745.png 1_26_26828.png 1_26_26980.png 1_27_27075.png 1_27_27775.png 1_28_28395.png 1_28_28547.png 1_29_29297.png 1_29_29762.png 1_29_29788.png 1_30_30065.png 1_31_31077.png 1_31_31675.png 1_32_32150.png 1_32_32195.png 1_32_32486.png 1_32_32606.png 1_32_32951.png 1_33_33625.png 1_34_34288.png 1_34_34382.png 1_34_34999.png 1_35_35222.png 1_35_35795.png 1_36_36237.png 1_36_36717.png 1_36_36781.png 1_37_37003.png 1_37_37335.png 1_37_37508.png 1_37_37791.png 1_38_38060.png 1_38_38093.png 1_38_38399.png 1_38_38642.png 1_38_38661.png 1_38_38763.png 1_39_39114.png 1_39_39391.png 1_39_39648.png 1_39_39670.png 1_39_39949.png 1_40_40818.png 1_41_41677.png 1_42_42429.png 1_42_42923.png 1_43_43257.png 1_43_43268.png 1_43_43730.png 1_44_44108.png 1_44_44118.png 1_45_45131.png 1_45_45191.png 1_45_45246.png 1_45_45286.png 1_45_45397.png 1_45_45762.png 1_45_45784.png 1_46_46108.png 1_46_46171.png 1_46_46443.png 1_46_46856.png 1_47_47122.png 1_47_47184.png 1_47_47611.png 1_47_47764.png 1_47_47779.png 1_47_47781.png 1_47_47818.png 1_47_47851.png 1_47_47854.png 1_47_47956.png 1_48_48173.png 1_48_48240.png 1_48_48451.png 1_48_48466.png 1_48_48572.png 1_48_48594.png 1_49_49284.png 1_49_49609.png 1_49_49612.png 1_49_49937.png 1_50_50473.png 1_50_50477.png 1_50_50516.png 1_50_50665.png 1_50_50979.png 1_51_51078.png 1_51_51123.png 1_51_51295.png 1_52_52303.png 1_53_53538.png 1_53_53692.png 1_53_53918.png 1_54_54098.png 1_54_54860.png 1_55_55005.png 1_55_55512.png 1_55_55530.png 1_55_55623.png 1_55_55762.png 1_55_55826.png 1_55_55915.png 1_55_55978.png 1_56_56060.png 1_56_56083.png 1_56_56172.png 1_56_56243.png 1_56_56265.png 1_56_56371.png 1_56_56546.png 1_57_57303.png 1_57_57848.png 1_57_57942.png 1_58_58095.png 1_58_58150.png 1_58_58478.png 1_58_58671.png 1_58_58820.png 1_58_58837.png 1_59_59160.png 1_59_59204.png 1_59_59208.png 1_59_59240.png 1_59_59300.png 1_59_59420.png 1_59_59466.png 1_59_59625.png 1_59_59743.png 1_60_60441.png 1_60_60473.png 1_60_60686.png 1_60_60982.png 1_61_61059.png 1_61_61147.png 1_61_61269.png 1_61_61280.png 1_61_61292.png 1_61_61579.png 1_61_61845.png 1_61_61911.png 1_62_62467.png 1_62_62514.png 1_62_62852.png 1_63_63034.png 1_63_63372.png 1_63_63410.png 1_63_63567.png 1_63_63633.png 1_63_63720.png 1_64_64723.png 1_64_64878.png 1_64_64938.png 1_65_65241.png 1_65_65815.png 1_65_65819.png 1_65_65838.png 1_65_65856.png 1_65_65984.png 1_66_66053.png 1_66_66598.png 1_66_66889.png 1_66_66903.png 1_66_66979.png 1_66_66988.png 1_67_67033.png 1_67_67050.png 1_67_67163.png 1_67_67202.png 1_67_67219.png 2_67_67381.png 2_67_67412.png 2_67_67426.png 2_67_67911.png 2_67_67978.png 2_68_68009.png 2_68_68315.png 2_68_68401.png 2_68_68614.png 2_68_68615.png 2_68_68619.png 2_68_68620.png 2_68_68707.png 2_68_68857.png 2_69_69224.png 2_69_69370.png 2_69_69372.png 2_69_69494.png 2_69_69499.png 2_69_69612.png 2_69_69648.png 2_69_69781.png 2_69_69989.png 2_70_70304.png 2_70_70531.png 2_70_70577.png 2_70_70612.png 2_70_70640.png 2_70_70803.png 2_70_70895.png 2_71_71115.png 2_71_71150.png 2_71_71279.png 2_71_71449.png 2_71_71768.png 2_71_71951.png 2_72_72074.png 2_72_72102.png 2_72_72562.png 2_72_72593.png 2_72_72714.png 2_73_73091.png 2_73_73097.png 2_73_73123.png 2_73_73629.png 2_73_73776.png 2_73_73838.png 2_74_74023.png 2_74_74056.png 2_74_74102.png 2_74_74166.png 2_74_74171.png 2_74_74294.png 2_74_74361.png 2_74_74560.png 2_74_74953.png 2_75_75089.png 2_75_75169.png 2_75_75273.png 2_75_75278.png 2_75_75430.png 2_75_75568.png 2_75_75665.png 2_75_75707.png 2_75_75751.png 2_75_75986.png 2_75_75991.png 2_76_76096.png 2_76_76149.png 2_76_76256.png 2_76_76257.png 2_76_76273.png 2_76_76631.png 2_76_76635.png 2_76_76639.png 2_76_76889.png 2_76_76940.png 2_76_76952.png 2_76_76961.png 2_77_77153.png 2_77_77162.png 2_77_77181.png 2_77_77435.png 2_77_77513.png 2_77_77805.png 2_78_78416.png 2_78_78606.png 2_78_78641.png 2_78_78701.png 2_78_78784.png 2_78_78817.png 2_78_78867.png 2_79_79244.png 2_79_79248.png 2_79_79319.png 2_79_79323.png 2_79_79407.png 2_79_79584.png 2_79_79799.png 2_80_80209.png 2_80_80262.png 2_80_80272.png 2_80_80334.png 2_80_80504.png 2_80_80562.png 2_80_80595.png 2_80_80829.png 2_80_80834.png 2_81_81295.png 2_81_81442.png 2_81_81675.png 2_81_81676.png 2_81_81973.png 2_81_81996.png 2_82_82045.png 2_82_82074.png 2_82_82348.png 2_82_82372.png 2_82_82411.png 2_83_83048.png 2_83_83262.png 2_83_83493.png 2_83_83591.png 2_83_83593.png 2_83_83604.png 2_83_83786.png 2_83_83792.png 2_83_83872.png 2_83_83877.png 2_84_84202.png 2_84_84204.png 2_84_84233.png 2_84_84392.png 2_84_84564.png 2_85_85029.png 2_85_85032.png 2_85_85416.png 2_85_85878.png 2_85_85925.png 2_85_85947.png 2_85_85964.png 2_86_86008.png 2_86_86011.png 2_86_86179.png 2_86_86180.png 2_86_86185.png 2_86_86190.png 2_86_86206.png 2_86_86299.png 2_86_86311.png 2_86_86425.png 2_86_86449.png 2_86_86516.png 2_86_86534.png 2_86_86575.png 2_86_86610.png 2_86_86832.png 2_86_86904.png 2_86_86945.png 2_87_87675.png 2_87_87881.png 2_88_88168.png 2_88_88294.png 2_88_88618.png 2_89_89079.png 2_89_89090.png 2_89_89100.png 2_89_89210.png 2_89_89214.png 2_89_89272.png 2_89_89273.png 2_89_89281.png 2_89_89299.png 2_89_89301.png 2_89_89429.png 2_89_89464.png 2_89_89493.png 2_89_89895.png 2_89_89940.png 2_89_89952.png 2_89_89954.png 2_90_90079.png 2_90_90135.png 2_90_90354.png 2_90_90390.png 2_90_90674.png 2_90_90726.png 2_90_90859.png 2_91_91016.png 2_91_91042.png 2_91_91073.png 2_91_91116.png 2_91_91198.png 2_91_91457.png 2_91_91491.png 2_91_91603.png 2_91_91670.png 2_92_92067.png 2_92_92133.png 2_92_92306.png 2_92_92312.png 2_92_92683.png 2_92_92712.png 2_92_92937.png 2_93_93142.png 2_93_93275.png 2_93_93311.png 2_93_93335.png 2_93_93399.png 2_93_93686.png 2_93_93790.png 2_93_93961.png 2_93_93981.png 2_93_93988.png 2_94_94130.png 3_100_100074.png 3_100_100139.png 3_100_100159.png 3_100_100170.png 3_100_100376.png 3_100_100465.png 3_100_100517.png 3_100_100631.png 3_100_100791.png 3_100_100991.png 3_101_101034.png 3_101_101035.png 3_101_101060.png 3_101_101115.png 3_101_101415.png 3_1

[krz02]

Nathan, i'd like to ask. what does max F beta score means ? it means highest f score from all of dataset? example: highest f score from ECSSD dataset. Or just calculating f beta score from each images , then averaged ?

[xuebinqin]

For example, given ECSSD dataset,

1. binarize each prediction in ECSSD with thresholds [0,255] to generate 256 binarize masks.

2. for each threshold compute the average F measure and get a average so that for 256 thresholds you will get 256 average F measure.

3. Find the max from the 256 average F measures.

The exact implementation might be a bit different but the general computation idea is that same.

On Mon, Oct 21, 2019 at 2:02 AM evv21 notifications@github.com wrote:

Nathan, i'd like to ask. what does max F beta score means ? it means highest f score from all of dataset? example: highest f score from ECSSD dataset. Or just calculating f beta score from each images , then averaged ?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/19?email_source=notifications&email_token=ADSGORN4AAZYDV4BXHF6GRLQPVOY7A5CNFSM4I66AUQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBZNQPA#issuecomment-544397372, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORIKQM52EOJD5ISMQXDQPVOY7ANCNFSM4I66AUQQ .

-- Xuebin Qin PhD Candidate Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/