Leveraging LLM Embeddings for Cross Dataset Label Alignment and Zero Shot Music Emotion Prediction
This addresses the problem of disjoint emotion labels across music datasets for researchers and practitioners in music information retrieval, though it appears incremental as it builds on existing LLM and MERT methods.
The paper tackles music emotion recognition by using LLM embeddings to align labels across datasets and enable zero-shot prediction on novel emotion categories, achieving generalization to unseen labels without additional training.
In this work, we present a novel method for music emotion recognition that leverages Large Language Model (LLM) embeddings for label alignment across multiple datasets and zero-shot prediction on novel categories. First, we compute LLM embeddings for emotion labels and apply non-parametric clustering to group similar labels, across multiple datasets containing disjoint labels. We use these cluster centers to map music features (MERT) to the LLM embedding space. To further enhance the model, we introduce an alignment regularization that enables dissociation of MERT embeddings from different clusters. This further enhances the model's ability to better adaptation to unseen datasets. We demonstrate the effectiveness of our approach by performing zero-shot inference on a new dataset, showcasing its ability to generalize to unseen labels without additional training.