Xiong (Moon) Zhang

In fact, the results are not as accuracy as the TPAMI 2018 paper [18].

I trained the model with 8 SXM4 A100 GPUs (each with 80GB memory) with the configuration as listed in bevdet4d-sttiny.py, and my training log are as bellow, Formating bboxes of...

> @MandyMo Thanks for you infos! The lower result is derived from that the parameter trans_only is set True in eval pipeline. I will fixed this soon. You just need...

Thank you, I am fine. The model.zip is the SMPL model, except that we re-write the shape and pose components. Anything else what can I do for you ?