martinwholtmon / IT3920-2024-Master-MSIT

Master project for MSIT 2024 - Towards Efficient Human Action Recognition: The Role of Keyframe Selection in Video Processing
MIT License
0 stars 0 forks source link

Apply the padding during collate before batching instead of during the sampling stage #61

Closed martinwholtmon closed 2 months ago

martinwholtmon commented 3 months ago

When sampling frames, instead of padding it directly there, instead do it during the collate to batch. This might lower the computational cost as less frames are sometimes passed through the transformations.

Should probably verify if this is a good idea or not, as the last few inputs would be static.