TY - RPRT TI - An interpretable speech foundation model for depression detection by revealing prediction-relevant acoustic features from long speech AU - Qingkun Deng AU - Saturnino Luz AU - Sofia de la Fuente Garcia PY - 2026 DO - 10.21437/interspeech.2025-1968 UR - https://arxiv.org/abs/2406.03138 ID - 2406.03138 ER -