Abstract: Facial Expression Recognition (FER) is a crucial task in affective computing, but its conventional focus on the seven basic emotions limits its applicability to the complex and expanding ...
Abstract: We present a modular pipeline for summarizing broadcast news videos using large language and vision models, specifically integrating Whisper for ASR, TransNetV2 for shot segmentation, LLaVA ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results