BVPEC: A Cross-Modal BERT-ViT Framework for Performance Emotion Recognition from Multimodal Acting Data. IJCAI 2025, 49 (21). https://doi.org/10.31449/inf.v49i21.10434.