Abstract: Depression detection using multimodal signals has garnered growing attention due to its potential for early warning. However, existing approaches often rely on limited visual features or ...
Abstract: Hyperspectral anomaly detection (HAD) faces a significant challenge in separating scarce, small and subtle anomalous targets from complex backgrounds. For typical HAD approaches, background ...