Yang Yue
Papers
2
Total Citations
6
H-Index
2
About
Yang Yue is a rising researcher at the forefront of embodied AI and multimodal learning, with a sharp focus on bridging large language models with real-world robotic systems. His most impactful work, "DeeR-VLA: Dynamic Inference of Multimodal Large Language Models for Efficient Robot Execution" (2024), tackles a critical bottleneck in robotics: the computational cost of deploying multimodal large language models (MLLMs) on physical robots. Yue’s key contribution is a dynamic inference framework that enables MLLMs to selectively activate only the necessary visual and language processing modules during task execution—dramatically reducing latency and energy consumption without sacrificing comprehension or reasoning. This innovation is pivotal for making generalist robotic MLLMs practical for complex, real-time human-robot interaction. With early citations already accumulating (over 6 across two versions), his work is rapidly gaining traction in the robotics and AI communities. Yue’s research exemplifies the next wave of efficient, context-aware embodied intelligence, positioning him as a promising voice in the quest to build robots that truly understand and act on human instructions.
Research Focus
Key Achievements
Top Papers
- 1
- 2