Leonardo Chang

Advanced Technologies Application Center, Tecnológico de Monterrey

Papers

Total Citations

H-Index

About

Leonardo Chang’s research bridges computer vision, embedded systems, and medical robotics, with a focus on real-time image analysis for surgical applications. His early work introduced an FPGA-based approach for detecting SIFT interest keypoints (33 citations), demonstrating how hardware acceleration can enable computationally intensive vision algorithms on resource-constrained devices. More recently, Chang has advanced the field of medical instrument segmentation in endoscopic procedures. He led the development of YOLACT++-based systems that achieve real-time, robust instance segmentation of laparoscopic tools, directly addressing challenges posed by the ROBUST-MIS Challenge. His 2021 paper on assessing YOLACT++ for this task (24 citations) highlights its potential to improve patient safety by enabling precise, low-latency tracking during computer- and robotic-assisted surgeries. A follow-up work introduced an attention-enhanced variant (2 citations), further refining segmentation accuracy. By combining algorithmic innovation with practical deployment considerations, Chang’s contributions support the next generation of intelligent surgical systems, where reliable visual feedback is critical for both autonomous and assisted procedures.