Overview
PUSHVIC is at the forefront of research in Augmented Reality (AR), Robotics, and Computer Vision. We explore spatial computing and human visual perception to bridge the fundamental research to real-world deployment. Our work advances AR, robotics, and vision-based systems to enhance human interaction, perception, and decision-making. Through interdisciplinary research and industry collaboration, we develop technologies that expand the frontiers of visual and spatial computing.
Technical Focus
Our core research lies in vision-language integration, robust 3D scene understanding, and visual localization.
Key areas includes:
- 3D Scene Understanding
- Robust Camera Relocalization
- Fast and Accurate 3D Reconstruction
- Vision-Language Models (VLMs) for Spatial Reasoning
- Failure-Robust HD Maps for Autonomous Driving
- Neural Rendering (e.g., Gaussian Splatting from Images)
