Chenyang LEI
Chenyang LEI
应该是相机的white level和black level，具体含义可以去搜索一下。
We did not try this before. I am not sure about the performance. But image inpainting is a little different from other image to image tasks because many pixels in...
We directly use the training data in this repo https://github.com/phoenix104104/fast_blind_video_consistency so we do not have the script
Hi, sorry for the mistake. The names are wrong for these two images. Yes, the reflection of the flash should be strong.
SURE. Thank you very much.
Sorry, we do not conduct this experiment.
It should be correct. The "_infer" suffix is used as "--training_mode test ".
Sorry, I am not familiar with video flicker detection. We assume human can tell what is a flickering video. I believe video flicker detection would be an interesting topic to...
Apologies. I believe that is because there are different scenes in a single video. You need to split them to different videos and generate process them one by one.
Currently the maximum number is 200. You might need to split the videos to small clips and then use our algorithm for better performance.