Junyi Yang

Papers

Total Citations

H-Index

About

Dr. Junyi Yang is a leading researcher in computer vision and marine artificial intelligence, with a focus on robust object detection in challenging underwater environments. Her most impactful work, "YOLO-based marine organism detection using two-terminal attention mechanism and difficult-sample resampling" (2024, 44 citations), addresses critical challenges in underwater imaging: noise interference and severe class imbalance in marine-life datasets. Dr. Yang introduced a novel two-terminal attention mechanism that enhances feature extraction from degraded images, combined with a difficult-sample resampling strategy that improves detection of rare or underrepresented species. This work has significantly advanced automated marine biodiversity monitoring, enabling more accurate and reliable identification of organisms in real-world conditions. Her contributions are particularly valuable for ecological conservation and fisheries management, where traditional detection methods often fail. With her innovative approach to integrating attention mechanisms into the YOLO architecture, Dr. Yang has set a new standard for object detection in noisy, imbalanced datasets, earning recognition as a rising authority in the intersection of deep learning and marine science.