Yehui Shen

National University of Defense Technology, Northeastern University, Southwest University of Science and Technology

Papers

Total Citations

H-Index

About

Yehui Shen is an emerging researcher specializing in autonomous robotics, visual place recognition (VPR), and simultaneous localization and mapping (SLAM). Their work addresses some of the most pressing challenges in robot navigation, particularly enabling reliable localization across diverse and dynamically changing environments. Shen's most notable contributions center on advancing VPR through innovative deep learning frameworks. Their 2025 paper introducing UGNA-VPR pioneered a training paradigm leveraging uncertainty-guided Neural Radiance Field (NeRF) augmentation to overcome the limitations of single-viewpoint datasets, earning 3 citations shortly after publication. Complementing this, their 2024 work on the Teacher-Student Cross-Metric Knowledge Distillation model (TSCM) demonstrated a sophisticated approach to making VPR systems more robust against lighting and weather variations. Shen also contributed meaningfully to LiDAR-based SLAM through semantic range image techniques for improved loop closure detection, reflecting a versatile grasp of both vision and point-cloud sensing modalities. Branching beyond navigation software, their 2025 work on spinal bistable oscillators for crawling robots signals a growing interest in bio-inspired mechanical systems. Though early in their career, Shen's interdisciplinary contributions across perception, localization, and robotics hardware position them as a promising voice in autonomous systems research.