MichiganCOG / ViP

Video Platform for Action Recognition and Object Detection in Pytorch
MIT License
220 stars 37 forks source link

Translation preprocessing function #28

Closed natlouis closed 5 years ago

natlouis commented 5 years ago

Referencing ticket #27 for both bounding boxes and coordinate points.

The same transform is used for all frames in a video, but the translation amount is bounding by the object's position. Meaning, the object should always be in view (never cropped out).