Language: English | 繁體中文

ASSURED COMPUTER VISION — LEAN, AUTONOMOUS, BROAD-SPECTRUM

Dr. Chih-Chung Hsu (許志仲)

Associate Professor, IEEE Senior Member

NYCU Young Chair Professorships
Institute of Smart Industry and Green Energy, College of Artificial Intelligence
National Yang Ming Chiao Tung University

Adjunct Associate Professor
Institute of Data Science, National Cheng Kung University

The Advanced Computer Vision Lab (ACVLab) builds deployment-grade visual intelligence organized around four research pillars. Assured Visual Intelligence ensures that AI outputs can be trusted in forensic, medical, and safety-critical settings — from detecting AI-generated content in social-media streams to proactive media authentication. Lean Visual Architectures rethink computation at every abstraction level — bitstream-level analysis, prefix-scan attention, and quantization-aware optimization — so that models run on edge devices, satellites, and real-time autonomous systems. Autonomous Visual Perception extends vision into 3D space for material-aware scene reconstruction, self-driving defense, and embodied robotic agents. Broad-Spectrum Scientific Sensing pushes perception beyond RGB to hyperspectral satellite imagery for Earth observation, climate monitoring, and precision agriculture. These pillars interlock: hyperspectral forensics merges trust with spectrum; on-satellite compressed sensing merges efficiency with spectrum; and BEV adversarial defense merges trust with embodiment.

Contact
chihchung [at] nycu.edu.tw • +886-6-303-2121 ext. 57928
Assistant: Ms. Jean Lin (林君柔小姐) • jeanlin [at] nycu.edu.tw • +886-972-325-397