CustomDancer: Customized Dance Recommendation by Text-Dance Retrieval
For users seeking personalized dance recommendations, this work provides a novel retrieval method, though performance is still low (10.23% R@1), indicating incremental progress.
The paper tackles the underexplored problem of text-based dance retrieval by introducing a large-scale dataset (TD-Data) and a multimodal retrieval framework (CustomDancer), achieving state-of-the-art performance with 10.23% Recall@1.
Dance serves as both a cultural cornerstone and a medium for personal expression, yet the rapid growth of online dance content has made personalized discovery increasingly difficult. Text-based dance retrieval offers a natural interface for users to search with choreographic intent, but it remains underexplored because dance requires simultaneous reasoning over linguistic semantics, musical rhythm, and full-body motion dynamics. We introduce TD-Data, a large-scale open dataset for text-dance retrieval, containing about 4,000 12-second dance clips, 14.6 hours of motion, 22 genres, and annotations from professional dance experts. On top of this dataset, we propose CustomDancer, a multimodal retrieval framework that aligns text with dance through a CLIP-based text encoder, music and motion encoders, and a music-motion blending module. CustomDancer achieves state-of-the-art performance on TD-Data, reaching 10.23% Recall@1 and improving retrieval quality in both quantitative benchmarks and user preference studies.