Z-Z-J

Results 8 comments of Z-Z-J

I meet the same problem, have you solved it now?

> 我运行脚本 > ![image](https://user-images.githubusercontent.com/33227739/90630322-6b627900-e253-11ea-9d93-912316766744.png) I solved this problem . sudo apt install ffmpeg

@NuochengTian Hi,you can provide the h36m's preprocessing script ? Thanks!!

![QQ截图20230331090410](https://user-images.githubusercontent.com/53818279/228997052-928851a7-44d6-4b14-ae87-2e1fdb0b851b.png) when 3D pose GT is taken as input, in Tab.4 the performance: 29.0 MPJPE and 23.0 PA-MPJPE on H36M. But in Tab.13, the performance: 13.9MPJPE 9.9 PA-MPJPE on H36M....

I attempt to train MeshNet with GT pose as input on H36M (using SMPL skeleton). ![QQ截图20230331133227](https://user-images.githubusercontent.com/53818279/229031822-8137a7a6-eba3-4fef-bf5c-eae0ac2450a9.png) ![QQ截图20230331133244](https://user-images.githubusercontent.com/53818279/229031825-e06e5d8c-24c2-4ec6-abe9-eb908804dfa9.png) In training set, we observe that MPVE: 6.3 mm, MPJPE: 11.9mm But in...

Thank you for your reply! Can you provide the pre-training model (only training MeshNet with 3D GT)?

In the process of processing data (run.py line: 114), they remove global offset, but keep trajectory in first position. Because in previous works, such as Videopose3d, these works need to...

> > > > Hello and thanks for the answer. Do you also know why they aren't excluding this output completely? Wouldn't it be helpful to estimate the 3D keypoints...