Cross-Modal Transformer With Dynamic Attention Fusion for Emotion Recognition in Music via Audio-Lyrics Alignment. IJCAI 2025, 49 (28). https://doi.org/10.31449/inf.v49i28.11516.