ROMP
ROMP copied to clipboard
Arthur151
Monocular, One-stage, Regression of Multiple 3D People and their 3D positions & trajectories in camera & global coordinates. ROMP[ICCV21], BEV[CVPR22], TRACE[CVPR2023]
感谢您之前的指导，我已经可以根据单人视频的参数在unity中驱动单人模型了~ 对于人数不变的视频，我发现single_batch_results每帧输出的人物的参数的数组是对应的，第一帧的single_batch_results[0]和第二帧的single_batch_results[0]是同一个人。 但是对于人数变化的视频，请问我应该怎么知道每帧视频中人物是如何对应的呢~
大佬您好，非常感谢您的工作~ 我在复现论文结果的过程中遇到了一些问题 我采用了v1.yml来进行训练，由于显卡原因（我的电脑一张3070显卡，8G显存）我的batch调小到了8，其他基本都是默认，以下是我的yml文件 [v1.yml.txt](https://github.com/Arthur151/ROMP/files/9052026/v1.yml.txt) 我训练了5个epoch之后，发现Loss虽然在下降，但是下降的很慢，下面是我的log [v1_hrnet-0704-002.log](https://github.com/Arthur151/ROMP/files/9052034/v1_hrnet-0704-002.log) 我之前有尝试过将3DPW的数据集加入到训练之中，发现很快就达到了不错的结果，下面是加入了3DPW之后的log [v1_hrnet-0704-001.log](https://github.com/Arthur151/ROMP/files/9052044/v1_hrnet-0704-001.log) 我想确认下我的训练是否有问题，因为看起来虽然下降的比较慢，但是还是在下降的 1、想请大佬帮忙看下日志，能否确认我的训练是有问题的状态？ 2、大佬能否发下您的训练日志（使用了pretrian/pretrain_hrnet.pkl），然后我自己对比一下 PS：我没有使用最新的代码，因为看起来和代码并没有关系
Thank your works! Your thesis is very impressive and wonderful. I have 2 Questions. I understand that the dataset below is used when training BEVs. dataset: h36m,muco,coco,lsp,agora,pw3d,relative_human,posetrack eval_datasets: 'relative,agora' However,...
Hi, long time no see. Glad to see your BEV's contribution. I have been playing with ROMP+blender now and developing something interesting. But the translation along the depth direction is...
Hi, I got another bug when I tried to evaluate the model. Here is my command ``` CUDA_VISIBLE_DEVICES=0 python romp/lib/evaluation/collect_3DPW_results.py --configs_yml=configs/eval_3dpw_challenge.yml ``` Here is the bug report ``` Traceback (most...
作者你好，我使用.npz文件转.fbx的时候，设置了fps_source，fps_target这两个参数为30，但转换后的fbx结果帧率还是24，如图所示  请问这是什么问题
Hi, Doc. Sun. It's me again. I find eval_AGORA.py still contains some bugs. When I run this script. I find the outputs['joints'] and the outputs['pj2d_org'] are ndarrays, but would be...
Hello! I have recently run your work. I want to get one main person from my video. However there are some other persons showing up during the video sometimes. How...
Hi, Really impressive work with ROMP and BEV! I was wondering generally around both implementations - why do you decide to create a dense parameter map for things like smpl...
[According to the README](https://github.com/Arthur151/ROMP/tree/master/simple_romp#calling-as-python-lib) i need to use the following code for using it as an module in 'image' mode: ```python #!/usr/bin/env python3 import romp settings = romp.main.default_settings romp_model =...
