Official PyTorch implementation for ICCV2023 paper "Exploring Predicate Visual Context in Detecting Human-Object Interactions"
BSD 3-Clause "New" or "Revised" License
56
stars
7
forks
source link
add options to obtain object features via `RoIAlign` #25
Closed
fredzzhang closed 11 months ago