CGS-BR

CGS-BR is the first publicly released visual dataset for detecting respiratory behaviors in the Chinese giant salamander (Andrias davidianus), addressing the critical lack of annotated data for rare amphibian monitoring. Comprising 1,632 high-definition images extracted from 215 video clips under controlled aquaculture conditions, the dataset covers diverse lighting and postures, with every respiratory cycle meticulously labeled frame-by-frame into four phases: HeadUP, Dive, Exhale, and Inhale. As a foundational resource for deep learning-based health assessment, conservation AI, and early disease detection, CGS-BR enables robust behavior analysis for this critically endangered species.Researchers studying the giant salamander are encouraged to contact the corresponding author at mailto:[e-mail] to sign a data use agreement and obtain download permissions.