Shijie Li

Papers

Total Citations

H-Index

About

Shijie Li is an emerging researcher working at the intersection of computer vision, deep learning, and robotic perception, with a particular focus on RGB-D semantic understanding and synthetic data generation. Li's most notable contribution explores the challenge of training data scarcity in semantic image segmentation for depth-sensing cameras — a problem especially critical when robots operate in privacy-sensitive environments such as private homes, where large-scale data collection is inherently restricted. To address this limitation, Li's work on "Semantic RGB-D Image Synthesis" (2023) proposes innovative approaches to generating diverse, annotated training images synthetically, reducing dependence on real-world data collection while maintaining segmentation performance. This research tackles a fundamental bottleneck in robotics and autonomous systems: enabling machines to understand their environments reliably even when training data is limited or difficult to obtain ethically. Though early in citation trajectory with 2 citations, the work addresses timely and high-impact challenges as privacy concerns and data accessibility increasingly shape the development of intelligent systems. Li represents a promising voice in the growing field of synthetic data-driven robot perception.