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
Is your feature request related to a problem? Please describe.
Mehul mentioned that it would be useful for behaviour data that is synced to neural recordings to accommodate variable frame rate.
Describe the solution you'd like
One way could be similar to the bounding boxes approach.
The sample dataset from MOCA could be seen as a variable frame rate dataset (the labels are every 5 frames, plus the last frame), and in PR #313 I relaxed the requirement for the extracted frames to just be monotonically increasing (and not necessarily equally spaced).
I am not sure how we do this for poses but maybe we could match them.
Describe alternatives you've considered
\
Additional context
\
The text was updated successfully, but these errors were encountered:
Is your feature request related to a problem? Please describe.
Mehul mentioned that it would be useful for behaviour data that is synced to neural recordings to accommodate variable frame rate.
Describe the solution you'd like
One way could be similar to the bounding boxes approach.
The sample dataset from MOCA could be seen as a variable frame rate dataset (the labels are every 5 frames, plus the last frame), and in PR #313 I relaxed the requirement for the extracted frames to just be monotonically increasing (and not necessarily equally spaced).
I am not sure how we do this for poses but maybe we could match them.
Describe alternatives you've considered
\
Additional context
\
The text was updated successfully, but these errors were encountered: