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
I've noticed that most of the video loading during training uses the DecordInit method with decord.VideoReader to access video frames. However, upon comparing it with mmcv.VideoReader(), I found that mmcv reads videos faster. Additionally, the frame count read by these two methods is inconsistent, with decord sometimes reading fewer frames than mmcv. Therefore, I have two questions:
Why not use the faster mmcv for video reading?
Why is there a discrepancy in the number of frames read by these two methods?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
I've noticed that most of the video loading during training uses the
DecordInitmethod withdecord.VideoReaderto access video frames. However, upon comparing it withmmcv.VideoReader(), I found that mmcv reads videos faster. Additionally, the frame count read by these two methods is inconsistent, with decord sometimes reading fewer frames than mmcv. Therefore, I have two questions:Why not use the faster mmcv for video reading?
Why is there a discrepancy in the number of frames read by these two methods?
Beta Was this translation helpful? Give feedback.
All reactions