GSoC 2024 | Contributor Application - Neural 3D Capture and Rendering

[Harvey Cheng]

Hello mentors, My name is Li-Wei(Harvey) Cheng. I'm a CS master's student in National Cheng Kung University in Taiwan. My research topic is about target-driven navigation using SLAM to build a semantic map and navigate.

I found the "Neural 3D Capture and Rendering" topic very interesting because I know dense SLAM is the foundation of all the other applications like 3D reconstruction that can be used to capture our own environment scene. By constructing digital twins, we might overcome the reality gap issue encountered in sim2real navigation training methods through high-quality reconstruction. Moreover, being able to make the 3DGS-based scene reconstruction method easily available for users is an exciting idea. I'm thrilled to make this happen with OpenCV team.

The following is my proposal and Linkedin page. Thanks for the great opportunity! I am keenly looking forward to exchange ideas and get feedback and suggestion from the OpenCV team.