Pytorch implementation of PersonEnvironmentNet (PEN), CVPR 2024
PEN.py: Official implementation of PersonEnvironmentNet
TF2023train.py: Official training split
TF2023eval.py: Official testing split
Annotation formats:
Example:["10", "00001.jpg", "T", "1", "2351"]
"10" - Scene number 10
"00001.jpg" - Frame number
"T" - Name of the third-person camera
"1" - Name of the first-person camera
"2351" - IDs of all people in this frame
TODO: Add download link for TF2023 Temporary download link