The Song Describer Dataset: a Corpus of Audio Captions for Music-and-Language Evaluation
Published in
arXiv.org(2023)External Links:
TL
TL;DR
This work introduces the Song Describer dataset (SDD), a new crowdsourced corpus of high-quality audio-caption pairs, designed for the evaluation of music-and-language models, and benchmark popular models on three key music- and-language tasks.
Abstract
We introduce the Song Describer dataset (SDD), a new crowdsourced corpus of high-quality audio-caption pairs, designed for the evaluation of music-and-language models. The dataset consists of 1.1k human-written natural language descriptions of 706 music recordings, all publicly accessible and released under Creative Common licenses. To showcase the use of our dataset, we benchmark popular models on three key music-and-language tasks (music captioning, text-to-music generation and music-language retrieval). Our experiments highlight the importance of cross-dataset evaluation and offer insights into how researchers can use SDD to gain a broader understanding of model performance.
References36 items
1
Music Understanding LLaMA: Advancing Text-to-Music Generation with Question Answering and Captioning
2
LP-MusicCaps: LLM-Based Pseudo Music Captioning
3
Language-Guided Music Recommendation for Video via Prompt Analogies
4
Audio-Text Models Do Not Yet Leverage Natural Language
5