Skip to content

Black band at certain videos #3534

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
bjuncek opened this issue Mar 9, 2021 · 0 comments
Open

Black band at certain videos #3534

bjuncek opened this issue Mar 9, 2021 · 0 comments

Comments

@bjuncek
Copy link
Contributor

bjuncek commented Mar 9, 2021

🐛 Bug

There is an issue in memory alignment before SWS scale which causes an error in about 4% of the frames on my sample of random videos from various datasets and about 1% of them on the test videos in TV.

The issue most likely lies in memory alignment in our usage of av_image_fill_arrays as this works when hardcoded av_frame buffer is allocated. I'm investigating the fix in the meantime

To Reproduce

Open and try to visualize the first frame of vision/test/assets/videos/WUzgd7C1pWA.mp4 using any method that relies on video_reader backend

Expected behavior

A full frame with no artifacts is decoded

Environment

PyTorch version: 1.8.0a0+d555768
Is debug build: False
CUDA used to build PyTorch: 11.2
ROCM used to build PyTorch: N/A

OS: Ubuntu 18.04.5 LTS (x86_64)
GCC version: (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0
Clang version: Could not collect
CMake version: version 3.19.4

Python version: 3.8 (64-bit runtime)
Is CUDA available: True
CUDA runtime version: Could not collect
GPU models and configuration:
GPU 0: Quadro RTX 8000
GPU 1: Quadro RTX 8000

Nvidia driver version: 460.27.04
cuDNN version: Probably one of the following:
/usr/local/cuda-10.1/targets/x86_64-linux/lib/libcudnn.so.7
/usr/local/cuda-10.2.89/targets/x86_64-linux/lib/libcudnn.so.7
HIP runtime version: N/A
MIOpen runtime version: N/A

Versions of relevant libraries:
[pip3] numpy==1.20.0
[pip3] torch==1.8.0a0+unknown
[pip3] torchvision==0.9.0a0+8ee9092
[conda] blas                      1.0                         mkl
[conda] magma-cuda112             2.5.2                         1    pytorch
[conda] mkl                       2020.2                      256
[conda] mkl-include               2020.2                      256
[conda] mkl-service               2.3.0            py38he904b0f_0
[conda] mkl_fft                   1.2.0            py38h23d657b_0
[conda] mkl_random                1.1.1            py38h0573a6f_0
[conda] numpy                     1.20.0                   pypi_0    pypi
[conda] torch                     1.8.0a0+unknown          pypi_0    pypi
[conda] torchvision               0.9.0a0+8ee9092           dev_0    <develop>

Additional context

This issue arose in #2916 and I'm opening a new one to have the change of the code needed (using torch.max instead of torch.mean for video stream tests in tests/test_video.py. I'm hoping to merge that PR and continue to track the issue here.

cc @bjuncek

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant