DOVER
DOVER copied to clipboard
Different videos get the same scores
Following the instructions provided, I have conducted evaluations using DOVER/DOVER++ checkpoints on several videos:
python evaluate_one_video.py -v video_1.mp4 -f
python evaluate_one_video.py -v video_2.mp4 -f
...
However, all these videos receive exactly the same scores (both the aesthetic and technical scores). Would you happen to have any ideas about why this happens? Thanks a lot!
same issue encountered in evaluate_a_set_of_videos.py