zhihou7 / HOI-CL

Series of work (ECCV2020, CVPR2021, CVPR2021, ECCV2022) about Compositional Learning for Human-Object Interaction Exploration
https://sites.google.com/view/hoi-cl
MIT License
76 stars 11 forks source link

About number of objects in V-COCO #15

Closed daoyuan98 closed 2 years ago

daoyuan98 commented 2 years ago

Hi, Thank you for your great work!

I found in ATL and FCL, you mentioned that there are only two kinds of objects. I wonder what does it mean because when I look at the annotation of VCOCO, I found there are 50+ types of objects in the annotations.

Thank you!

zhihou7 commented 2 years ago

Hi, thanks for your interest,

V-COCO only considers objects and instruments. For example, V-COCO only includes “eat obj”, i.e. treat "eat apple" and "eat orange" as the same class. Therefore, there are only 29 HOIs ( include 5 pose action) on V-COCO. There are around 69 objects on V-COCO. However, V-COCO does not require recognizing the specific type of object.

daoyuan98 commented 2 years ago

Hi, Got it! Thank you for your reply!