search

microsoft / multiview-human-pose-estimation-pytorch

This is an official Pytorch implementation of "Cross View Fusion for 3D Human Pose Estimation, ICCV 2019".
MIT License
544 stars 89 forks source link

Question about live Cameras #8

Closed lisa676 closed 5 years ago

lisa676 commented 5 years ago

Hi dear Chunyu Wang, thanks for your such a nice work. I want to know that this work can be implemented for live cameras instead of images? I want to use 2 live cameras. Is it possible in your work? I read your work but still I'm confused about my problem.

CHUNYUWANG commented 5 years ago

The speed may be an issue when it is used for real time pose processing from live cameras.

lisa676 commented 5 years ago

Yeah I'm already thinking about it but main problem is that how to calibrate cameras within your repository? For cameras calibration we need another method which can be embedded in your repository later or we can calibrate cameras directly in your repository? I'm unable to find any piece of code which can handle live camera calibration that's why I'm asking this question.

CHUNYUWANG commented 5 years ago

It does not have the code for camera calibration. We assume the camera parameters are already known as is the case for the benchmark datasets.

lisa676 commented 5 years ago

Yeah but for real-time processing scenario will be different. Thanks for your replies.

KungZell commented 2 years ago

Yeah but for real-time processing scenario will be different. Thanks for your replies.

Hello, dear Lisa, Recently I wanted to realize real-time 3D tracking through two cameras. When browsing GitHub, I found that you are the pioneer in this field. Could you please give me some suggestions ? Thank you very much.