Zaitian Gongye

Papers

Total Citations

H-Index

About

Zaitian Gongye is a rising star in embodied AI, whose work sits at the critical intersection of legged robotics, vision-language models, and autonomous navigation. His most impactful contribution is **NaVILA** (2025, 13 citations), a groundbreaking Vision-Language-Action (VLA) model that enables legged robots to follow complex, long-horizon natural language commands—such as "move forward out of the room" or "stop in front of the red door"—without requiring pre-mapped routes. This work, developed at NVIDIA, represents a paradigm shift from traditional SLAM-based navigation to language-guided, zero-shot generalization in unstructured environments. Gongye also advanced trajectory prediction with his work on **hierarchically depicting vehicle trajectory with stability** (2025, 10 citations), addressing the challenge of maintaining robust path planning in complex, dynamic settings. His research directly tackles the "last mile" of robotics: bridging high-level human intent with low-level motor control. By demonstrating that robots can interpret nuanced spatial instructions and execute them with human-like agility, Gongye is laying the foundation for truly autonomous service robots. His early citation impact signals a researcher whose ideas are already shaping the next generation of intelligent, language-driven robotic systems.