You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi. I am trying to use the idea400 video-text pairs. However, the text is named after the actions, and the videos are by numbers.
How could I find the matched text using the video names (the number)? Maybe there is other meta file that I missed out?
The text was updated successfully, but these errors were encountered:
In our dataset, for each sequence, the filenames of the corresponding video and semantic label are consistent, differing only in the file extension. For instance, a video-text pair may have filenames like Act_cute_and_sitting_at_the_same_time_clip1.mp4 and Act_cute_and_sitting_at_the_same_time_clip1.txt, with only the file extension differing, making it easy to match the files based on the filename. This pattern applies to other modalities as well.
Hi. I am trying to use the idea400 video-text pairs. However, the text is named after the actions, and the videos are by numbers.
How could I find the matched text using the video names (the number)? Maybe there is other meta file that I missed out?
The text was updated successfully, but these errors were encountered: