Closed ctyeong closed 1 year ago
This might be an issue with Helios, since the example notebook generates images on my system (Mac). What platform are you running on?
@pranavraja99 @dariojavo, would you have any insight into this issue?
@amogh7joshi, I am running on Ubuntu18
Hi @ctyeong, can you please help us with the following information: 1) Was Helios correctly installed in your computer? 2) If so, please share the output of the previous step:
# Generate the data.
generator.generate(name = 'tomato_sample', num_images = 3)
We have tested under Ubuntu 18 and 20, so it should be working. Please help us with that information in order to find out what is happening.
Here is the output from
generator.generate(name = 'tomato_sample2', num_images = 3)
Loading XML file: /home/username/anaconda3/envs/agml/lib/python3.7/site-packages/agml/_helios/Helios/projects/SyntheticImageAnnotation/xml/style_tomato_sample2.xml...done.
Reading XML file: /home/username/anaconda3/envs/agml/lib/python3.7/site-packages/agml/_helios/Helios/projects/SyntheticImageAnnotation/xml/style_tomato_sample2.xml...Building canopy of tomato plants...done.
Ground geometry...done.
Ground consists of 100 total primitives.
done.
/home/username/.agml/synthetic/tomato_sample2/image0Rendering RGB image containing 84.82K primitives...Initializing graphics...done.
writing JPEG image: /home/username/.agml/synthetic/tomato_sample2/image0/view00000/RGB_rendering.jpeg
writing JPEG image: /home/username/.agml/synthetic/tomato_sample2/image0/view00001/RGB_rendering.jpeg
writing JPEG image: /home/username/.agml/synthetic/tomato_sample2/image0/view00002/RGB_rendering.jpeg
done.
Generating labeled image containing 2 label groups...
Initializing graphics...done.
Performing semantic segmentation for view 0... and element:
clusters
leaves
Semantic segmentation ... done.
Performing semantic segmentation for view 1... and element:
clusters
leaves
Semantic segmentation ... done.
Performing semantic segmentation for view 2... and element:
clusters
leaves
Semantic segmentation ... done.
Loading XML file: /home/username/anaconda3/envs/agml/lib/python3.7/site-packages/agml/_helios/Helios/projects/SyntheticImageAnnotation/xml/style_tomato_sample2.xml...done.
Reading XML file: /home/username/anaconda3/envs/agml/lib/python3.7/site-packages/agml/_helios/Helios/projects/SyntheticImageAnnotation/xml/style_tomato_sample2.xml...Building canopy of tomato plants...done.
Ground geometry...done.
Ground consists of 100 total primitives.
done.
/home/username/.agml/synthetic/tomato_sample2/image1Rendering RGB image containing 84.484K primitives...Initializing graphics...done.
writing JPEG image: /home/username/.agml/synthetic/tomato_sample2/image1/view00000/RGB_rendering.jpeg
writing JPEG image: /home/username/.agml/synthetic/tomato_sample2/image1/view00001/RGB_rendering.jpeg
writing JPEG image: /home/username/.agml/synthetic/tomato_sample2/image1/view00002/RGB_rendering.jpeg
done.
Generating labeled image containing 2 label groups...
Initializing graphics...done.
Performing semantic segmentation for view 0... and element:
clusters
leaves
Semantic segmentation ... done.
Performing semantic segmentation for view 1... and element:
clusters
leaves
Semantic segmentation ... done.
Performing semantic segmentation for view 2... and element:
clusters
leaves
Semantic segmentation ... done.
Loading XML file: /home/username/anaconda3/envs/agml/lib/python3.7/site-packages/agml/_helios/Helios/projects/SyntheticImageAnnotation/xml/style_tomato_sample2.xml...done.
Reading XML file: /home/username/anaconda3/envs/agml/lib/python3.7/site-packages/agml/_helios/Helios/projects/SyntheticImageAnnotation/xml/style_tomato_sample2.xml...Building canopy of tomato plants...done.
Ground geometry...done.
Ground consists of 100 total primitives.
done.
/home/username/.agml/synthetic/tomato_sample2/image2Rendering RGB image containing 82.02K primitives...Initializing graphics...done.
writing JPEG image: /home/username/.agml/synthetic/tomato_sample2/image2/view00000/RGB_rendering.jpeg
writing JPEG image: /home/username/.agml/synthetic/tomato_sample2/image2/view00001/RGB_rendering.jpeg
writing JPEG image: /home/username/.agml/synthetic/tomato_sample2/image2/view00002/RGB_rendering.jpeg
done.
Generating labeled image containing 2 label groups...
Initializing graphics...done.
Performing semantic segmentation for view 0... and element:
clusters
leaves
Semantic segmentation ... done.
Performing semantic segmentation for view 1... and element:
clusters
leaves
Semantic segmentation ... done.
Performing semantic segmentation for view 2... and element:
clusters
leaves
Semantic segmentation ... done.
Moving Images: 100%|███████████████████| 9/9 [00:00<00:00, 13282.45it/s]
Seems that images are being generated correctly. Just to verify, for getting the loader you are using tomato_sample2
, right? Since you are generating the dataset with that name.
loader = agml.data.AgMLDataLoader.helios('tomato_sample2')
@dariojavo, exactly. Of course, the result won't change with tomato_sample
.
Yeah, of course, it should only work with the dataset that you generated. In this case tomato_sample2
.
Ok, lets check the images. Can you please confirm if the images on /home/username/.agml/synthetic/tomato_sample2/...
(e.g., /image2/view00000/RGB_rendering.jpeg
) are not empty?
image0-view00000.jpeg, image0-view00001.jpeg, image0-view00002.jpeg, image1-view00000.jpeg, image1-view00001.jpeg, image1-view00002.jpeg, image2-view00000.jpeg, image2-view00001.jpeg, image2-view00002.jpeg.
These all simply look black
Can you try running generator.generate
with convert_data = False
, and then check whether the output images have been correctly rendered?
I got an error below:
FileNotFoundError: The information file at '/home/username/.agml/synthetic/tomato_sample/.metadata/agml_info.json' for the Helios dataset tomato_sample could not be found.
Under ~/.agml.synthetic/tomato_sample/
, there are image0/view00000/RGB_rendering.jpeg
,image0/view00001/RGB_rendering.jpeg
... `image2/view00002/RGB_rendering.jpeg. But each one is still a black image.
Could the issue be related to the file path assuming /home/username
?
On Thu, Oct 6, 2022 at 8:58 AM Taeyeong Choi @.***> wrote:
I got an error below:
FileNotFoundError: The information file at '/home/username/.agml/synthetic/tomato_sample/.metadata/agml_info.json' for the Helios dataset tomato_sample could not be found.
Under ~/.agml.synthetic/tomato_sample/, there are image0/view00000/RGB_rendering.jpeg,image0/view00001/RGB_rendering.jpeg ... `image2/view00002/RGB_rendering.jpeg. But each one is still a black image.
— Reply to this email directly, view it on GitHub https://github.com/Project-AgML/AgML/issues/30#issuecomment-1270322104, or unsubscribe https://github.com/notifications/unsubscribe-auth/AG5FGPGB4V74YSOHCGQNNETWB3ZLRANCNFSM6AAAAAAQ5HBFPE . You are receiving this because you are subscribed to this thread.Message ID: @.***>
@masonearles I simply posted with username
here, which is not my actual username on my end.
@amogh7joshi @dariojavo Has this issue been resolved?
@masonearles @amogh7joshi I have checked and the error is on Helios when compiling in Ubuntu 18. I just reported this issue with Helios team.
Will keep this open until solved it.
@dariojavo Do we have resolution on this issue?
Yes, this has been solved! Tested on Ubuntu 18 and 20.
In your
AgML-Synthetic.ipynb
, I ran the following lines:This shows a pair of images as below, where the right seems to be correct with a segmentation result while the left is wrong with a black image.
I can find that
loader[0][0]
is simply a zero matrix becausenp.all(loader[0][0]==0)
outputsTrue
. I have tried this multiple times, but I always have the issue.