Multi-Task Music Representation Learning from Multi-Label Embeddings
This work addresses a bottleneck in music retrieval systems by improving triplet selection efficiency, though it appears incremental as it builds on existing triplet loss methods.
The paper tackles the problem of triplet selection in music representation learning by using multi-tag annotations and Latent Semantic Indexing to estimate tag-relatedness for hard triplet selection, achieving improved performance on a multi-task scenario with four new large multi-tag annotations for the Million Song Dataset.
This paper presents a novel approach to music representation learning. Triplet loss based networks have become popular for representation learning in various multimedia retrieval domains. Yet, one of the most crucial parts of this approach is the appropriate selection of triplets, which is indispensable, considering that the number of possible triplets grows cubically. We present an approach to harness multi-tag annotations for triplet selection, by using Latent Semantic Indexing to project the tags onto a high-dimensional space. From this we estimate tag-relatedness to select hard triplets. The approach is evaluated in a multi-task scenario for which we introduce four large multi-tag annotations for the Million Song Dataset for the music properties genres, styles, moods, and themes.