Naoto Tsukamoto

The University of Tokyo

Papers

3

Total Citations

14

H-Index

3

About

Naoto Tsukamoto is a robotics researcher advancing the frontier of autonomous systems through the integration of vision-language models (VLMs) and human-robot interaction. His work centers on enabling mobile robots to perceive and act in dynamic, unstructured environments without prior mapping or extensive training. In his highly cited 2023 paper, Tsukamoto introduced a method for semantic scene difference detection during daily-life patrolling, leveraging pre-trained large-scale VLMs to identify environmental changes—a critical capability for domestic service robots. This approach outperformed traditional anomaly detection techniques by focusing on semantic understanding rather than pixel-level differences. Building on this, his 2024 work on reflex-based open-vocabulary navigation demonstrated how an omnidirectional camera paired with multiple VLMs allows robots to navigate and respond to natural language commands without SLAM or reinforcement learning, achieving robust performance in zero-shot scenarios. Tsukamoto also developed a chat-based system for teaching robot action instructions, bridging the gap between non-expert users and complex robotic control. With over 14 citations across his recent papers, Tsukamoto’s contributions are shaping a future where robots understand and adapt to human environments intuitively, making him a rising voice in embodied AI and service robotics.

Research Focus

Key Achievements

3
H-Index
3
Papers
14
Total Citations
5
Avg Citations/Paper
🏆 Most Cited Paper
Semantic Scene Difference Detection in Daily Life Patroling by Mobile Robots Using Pre-Trained Large-Scale Vision-Language Model
8 citations · 2023
📈 Most Prolific Year: 2024 (2 Papers)
🤝 Key Collaborators: 10
🏛 Institutions: The University of Tokyo

Top Papers

  1. 1
  2. 2
  3. 3

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 12 days ago