-
![image](https://user-images.githubusercontent.com/31682925/63644810-b2428f00-c723-11e9-9466-b82702c96145.png)
if I want to use smplh as the model type , where should I download the corresponding .pk…
-
首先感谢一下开源(跪)
我们试着从一批FBX动画文件中批量导出关键点信息转成SMPL格式,但在使用这个工具时遇到了点儿问题
![image](https://user-images.githubusercontent.com/24589615/210829876-e67790d0-2b6d-490a-93be-8c74279e8a0e.png)
左图是example中给出的随机关…
-
First, Thank you for sharing codes!!
I'd like to synthesize novel pose on extended model. Novel pose means like random smpl poses, e.g. aist++ datasets or AMASS. datasets.
There is a code to retr…
-
Hi, Davis,
You really did a wonderful job for both your paper and your codebase! Thanks for sharing the code, which is well-organized and well-annotated. I love it and I have learned a lot.
I have…
-
Hi, thanks for this great work about egocentric body pose estimation. I'm trying to fit SMPL body model to this dataset but I find the naming in the json file is very different from SMPL and SMPL-H mo…
qzane updated
7 months ago
-
本人用Pytorch复刻了一版,也增加了valid过程,结果发现,FID_k最好的是第21个epoch,为101,后面越训练,FID_k的值越大,波动比较大,也不知道什么原因。 loss值降到0.0002左右,基本就不收敛了,FID_k的值能达到7000多, FID_g的值25左右。
使用原作者tensorflow原代码,从头开始训练,loss值降到0.0001后,基本就不收敛了,FID_k的值…
-
Hello again,
Thanks for releasing annotations for different datasets. I am trying to project the SMPL model from HuMMan dataset to the image using the SMPL parameters and camera parameters provide …
-
I processed the data as mentioned in the README and get the following error when I run
python visualizations/poses2concepts.py --config ./configs_kit_aug/MLP.yaml --dataset_name kit --n_map 5…
-
Hi, @linjing7 , can you help me with the OpenGL error message? Thanks
```
python3 demo.py --gpu 0 --img_path input.png --output_folder output_my --decoder_setting wo_decoder --pretrained_model_pa…
-
When I run `train_zju_mocap_refine.sh`, an error occurs: The issue seems to arise when attempting to reshape the array, as its size does not match the expected shape of (6890, 3).
![image](https://gi…