search

fredzzhang / pvic

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

Timeline for upadating Readme.md #34

Closed ZhouGuangP closed 8 months ago

ZhouGuangP commented 9 months ago

I am puzzled by the difference between my reproduced results on v-coco and the results reported in the article. I use 2 4090 GPU with batchsize=32 for reproducing. image

fredzzhang commented 9 months ago

Hi @ZhouGuangP,

The V-COCO code went through some major restructuring after the submission. I might have broken something in that process. I'm still trying fix it.

In the meantime, were you able to reproduce the performance on HICO-DET?

Fred.

ZhouGuangP commented 9 months ago

Hi @ZhouGuangP,

The V-COCO code went through some major restructuring after the submission. I might have broken something in that process. I'm still trying fix it.

In the meantime, were you able to reproduce the performance on HICO-DET?

Fred.

Yes. I reproduced the performance on HICO-DET, it is next to the results reported in your paper. The performance is 35.44map on full,31.06map on rare 35.44 on none-rare. Thanks for your nice work.

winnerwu6 commented 8 months ago

Hello, when will your training and evaluation code on the V-COCO data set be updated? My recent work requires your code very much. I will be very grateful if you can update it! !

fredzzhang commented 8 months ago

The documentation has now been updated.

Fred.