Arthur151 / ROMP

Monocular, One-stage, Regression of Multiple 3D People and their 3D positions & trajectories in camera & global coordinates. ROMP[ICCV21], BEV[CVPR22], TRACE[CVPR2023]
https://www.yusun.work/
Apache License 2.0
1.34k stars 229 forks source link

Fail to track more than one person in each frame #502

Open FrankBu0616 opened 8 months ago

FrankBu0616 commented 8 months ago

Hi, thank you for the amazing work. When I test out the demo CUDA_VISIBLE_DEVICES=0 trace2 -i /path/to/trace_demo2 --subject_num=2, only one person was tracked in each frame. Are there some extra parameters required? Thanks!

image

FrankBu0616 commented 8 months ago

After looking at the output it seems to be a visualization issue? Seems that people are tracked properly, but visualization assumed it was one person.

Posedg commented 4 months ago

oh,you have been successed running this demo,i'm sorry that i have no solution about your problem, i'm new here,could you help me with a question. It's when i did the step which is about trace's environment.(TRACE's code for inference & evaluation) putin(trace) h@h:~/ROMP/simple_romp$ python setup_trace.py install and i get some message like this 2024-05-31 15-01-46 的屏幕截图 i try a lot of solution,but failed. Thank you for your valuable time,Looking forward to hearing from you.