Yukun Zhu

Google (United States), University of Jinan

Papers

Total Citations

H-Index

About

Yukun Zhu is a researcher at the forefront of machine perception and intelligent sensing, with key contributions spanning computer vision and wearable electronics. In computer vision, Zhu has advanced semantic and panoptic segmentation—the pixel-level classification and instance identification critical for autonomous driving and robotics. Their work on "Superpixel Transformers for Efficient Semantic Segmentation" (2023, 11 citations) introduces a novel architecture that balances computational efficiency with high-dimensional pixel classification, addressing a core bottleneck in real-world deployment. Zhu also contributed to the "Waymo Open Dataset: Panoramic Video Panoptic Segmentation" (2022, 3 citations), a benchmark dataset that has become a standard resource for autonomous driving research, enabling more robust multi-object scene understanding. Beyond vision, Zhu’s interdisciplinary work in "High-Sensitivity Triboelectric Pressure Sensor with Dual-Microcone Synergistic Enhancement" (2025, 1 citation) tackles the sensitivity-dynamic range trade-off in flexible sensors, with applications in wearable electronics, electronic skin, and human-machine interaction. This sensor innovation demonstrates Zhu’s ability to bridge algorithmic perception with physical sensing systems. With a growing citation footprint and contributions to both foundational datasets and efficient architectures, Yukun Zhu is shaping the future of intelligent systems that see and feel their environment.