Shenhan Quian

This PhD project focuses on advancing visual perception and spatial intelligence within the domains of computer vision and artificial intelligence. The goal is to develop algorithms and models that enable machines to interpret visual data and reason about spatial relationships in complex, real-world environments. Visual perception involves the extraction of semantic and geometric information from images or video, while spatial intelligence pertains to understanding and navigating spatial layouts, object configurations, and physical dynamics.

The research will explore topics such as 3D scene reconstruction, depth estimation, object localization, spatial reasoning, and mapping. State-of-the-art techniques in deep learning, neural rendering, and geometric learning will be employed and extended to create more robust and generalizable systems. Special emphasis will be placed on enabling AI models to handle partial, ambiguous, or occluded visual inputs and still make accurate spatial inferences.

Applications of the work include autonomous navigation, robotic manipulation, augmented reality, etc. The outcomes are expected to contribute to the development of more spatially aware AI systems, capable of interacting with their environments in a structured and intelligent manner.