Proposing a novel Affective Multimodal Transformer model, Video2Music, to generate music that matches video content in terms of emotion.
