Enric Moreu
Enric Moreu
It is working for me, you have to add the `%04d` so the parser recognize your frames: `bash make-opt-flow.sh my_folder/my_frames/frame_%04d.ppm my_folder/my_opticalflow_folder` If you want to manually extract the frames, you...
I think preprocessing is not required for ShanghaiTech since the images are smaller and more uniform. You can run the training step directly You'll probably have to apply some changes...
Ubuntu 18.04 with CUDA 11.0 and the Docker-GPU drivers installed.