-
## ❓ How to do something using detectron2
I've seen that with Detectron is possible to have a sort of 3D mask of the body.
Is it possible to obtain 3D skeletons starting from these masks?
I need t…
-
我们在对2D模型预测出来的关键点做归一化时，发现摄像头画面里人身上的画线消失了，想知道 2D的输出数据做完归一化需要经过coco_h36m转化成h36m数据集的2d pose吗？， 下面是我的代码
-
Hi, I was wondering if your model works on the wild videos of which lower body is not visible and estimated only upper body key points.
-
Great work, and the code is clear!
However,I got a issue with 3D heatmap normalization.In paper(page 6,just below formula.2), it says:"∑_(u,v,d)H_3d(u,v,d)=1",and I wonder why,because that could make…
-
Items to be included in May's newsletter- will be sent out around 9th May
-
## 🐛 Bug
### To Reproduce
When training large datasets, the speed is very slow. About ten times slower than pytorch.
But when we reduced the dataset size, the training speed became normal a…
-
I try to inference YOLOv3 by openvino, there is an ie "api.pyd" file in the interface_engine folder of python3.6 folder.
Any method shown in this thread,all done.
Please teach me the tip to solve th…
-
Hi, how to convert the generated 3d pose to bvh or fbx?
-
This issue tracks the status of the tests of the TUG demo on the `R1SN003` robot.
cc @mfussi66 @pattacini @elandini84 @ste93 @lornat75
-
### Is there an existing issue for this?
- [X] I have searched the existing issues
### Bug description
Training starts and appears to recognize GPU but never reaches first iteration and GPU l…