This research work explores new technologies and trends in the field of computer vision, focusing on real-time 3D human pose tracking using deep learning algorithms. The goal is to create a system that allows tracking the movement and position of humans in real time in a laboratory, using images from multiple fixed cameras and a mobile camera. During this research, the uses of new technologies such as virtual reality and the digital simulation of a physical environment are going to be explored and evaluated. The core concept is to be able to achieve high accuracy in the tracking of the person using only an RGB video stream, which can be modeled and analyzed by different software in real-time. Great importance was placed on the design of the entire system in a manner that it can be easily extended and modified in the future. This is going to lay the foundations for more extensive and detailed research of new forms of human-machine interaction, new ways of remote management of a physical environment, and new forms of robotics.
- Professor Oswald Lanz, for the guidance
- Emanuele Caruso, for many insight and help with algorithms
- Stack Overflow
Insert gif or link to demo
The installation is complicated, I will provide more insights for a proper installation later on
This file wants to be a starting point for the documentation and be a more detailed explanation of the code. I hope I'll be able to insert also a program that downloads everything together in order to build and have everything ready.
If you have any feedback, please reach out to us at [email protected]
- Human Pose detection
- Real-time
- Unreal Engine integration
- Virtual/Augmented reality ready
- Faster data exchange between sockets
What optimizations did you make in your code? E.g. refactors, performance improvements, accessibility
Single Client: Unreal Engine, Caffe, Openpose, NDI SDK, Pylon SDK, Sockets
For support, email [email protected]
Computer science student based in Bolzano. Computer Vision and Virtual reality, 3D world. If you want to find out more, check out my website