Jamendo-MT-QA: A Benchmark for Multi-Track Comparative Music Question Answering
This benchmark addresses the lack of systematic evaluation for comparative music understanding, which is a common real-world listening scenario, but the contribution is incremental as it extends an existing dataset.
The paper introduces Jamendo-MT-QA, a benchmark for multi-track comparative music question answering, consisting of 36,519 QA items over 12,173 track pairs. The benchmark evaluates models on comparative reasoning across tracks, with results showing that current audio-language models perform poorly, highlighting the need for further research.
Recent work on music question answering (Music-QA) has primarily focused on single-track understanding, where models answer questions about an individual audio clip using its tags, captions, or metadata. However, listeners often describe music in comparative terms, and existing benchmarks do not systematically evaluate reasoning across multiple tracks. Building on the Jamendo-QA dataset, we introduce Jamendo-MT-QA, a dataset and benchmark for multi-track comparative question answering. From Creative Commons-licensed tracks on Jamendo, we construct 36,519 comparative QA items over 12,173 track pairs, with each pair yielding three question types: yes/no, short-answer, and sentence-level questions. We describe an LLM-assisted pipeline for generating and filtering comparative questions, and benchmark representative audio-language models using both automatic metrics and LLM-as-a-Judge evaluation.