Dekun Wu

Papers

Total Citations

H-Index

About

Dekun Wu is a robotics researcher whose work bridges computer vision and natural language processing to enhance autonomous systems' environmental understanding. His primary research focuses on scene classification for indoor robotics, where he has pioneered innovative approaches using context-based word embeddings to improve recognition accuracy. In his most cited work, "Scene Classification in Indoor Environments for Robots using Context Based Word Embeddings" (2019, 15 citations), Wu addresses the critical challenge of classifying increasingly complex indoor scene datasets—a task where traditional computer vision techniques often fall short. By integrating semantic context from word embeddings, his method enables robots to better interpret their surroundings, achieving more robust performance in real-world applications. This contribution is particularly significant for service and domestic robots operating in cluttered, dynamic indoor spaces. Wu's work exemplifies the growing trend of multimodal learning, where linguistic cues augment visual data to overcome the limitations of pure image-based classification. With his research gaining traction in the robotics community, Dekun Wu is establishing himself as a thoughtful contributor to more intelligent, context-aware autonomous navigation.