-
Hi!! Thanks for sharing your codes.
I have seen your result in your paper about bi-seg. I tried to reproduce the result of bi-seg, however I only got 71%IOU（single scale）, have you **successfully** g…
lxtGH updated
5 years ago
-
hey guys
I don't have access to Titan XP, but I have access to 1080 GPU on the same machine, is there any way I could train the model?
-
1.使用如下指令训练：python train.py --batch_size=16 --use_gpu=True --init_model="./model_1000/" --checkpoint_path="./chkpnt/"
其中：/model_1000为文档中提供的预训练模型，其中并无conv6参数。
2. 文档中eval和infer中的model_path建议改为./chkp…
-
hello, I run it on pytorch1.0.1 python3.5, but it has some problem with import models. when I run python ade20k.py --download-dir ./datasets/ade, I got "ImportError: No module named 'utils'"
How shou…
-
Is there a lower input size model available?
Will it help with faster inference if the image size is reduced?
How do I go about training on lower resolution images, what are the changes required in …
-
Hello,
I wonder, what's the training time, both for PSPNet baseline and ICNet. I am running PSPNet training at the moment, it looks like it takes ~3 days. Does that sound right?
-
```
root@63a11f01a6d1:/work/paddlepaddle/icnet# python train.py --batch_size=4 --use_gpu=True --checkpoint_path="./chkpnt/"
----------- Configuration Arguments -----------
batch_size: 4
checkpoin…
-
https://github.com/hellochick/ICNet-tensorflow/blob/a8f0eba4b40e60b7c198f96a780bf7448be67841/train.py#L24
https://github.com/hellochick/ICNet-tensorflow/blob/a8f0eba4b40e60b7c198f96a780bf7448be6784…
-
Nice work! I wonder how much time you speed to process the images of Cityscapes. Thus we can compare it with ICNet.
-
hi,nice work ！
when i train my dataset with your code from scratch, i modify the ''model_weight='./model/cityscapes/icnet_scapes_train_30k_bnnomerge.npy'''in train.py to
'model_weight = '' ' so …