WavCaps: A ChatGPT-Assisted Weakly-Labelled Audio Captioning Dataset for Audio-Language Multimodal Research
About
The advancement of audio-language (AL) multimodal learning tasks has been significant in recent years. However, researchers face challenges due to the costly and time-consuming collection process of existing audio-language datasets, which are limited in size. To address this data scarcity issue, we introduce WavCaps, the first large-scale weakly-labelled audio captioning dataset, comprising approximately 400k audio clips with paired captions. We sourced audio clips and their raw descriptions from web sources and a sound event detection dataset. However, the online-harvested raw descriptions are highly noisy and unsuitable for direct use in tasks such as automated audio captioning. To overcome this issue, we propose a three-stage processing pipeline for filtering noisy data and generating high-quality captions, where ChatGPT, a large language model, is leveraged to filter and transform raw descriptions automatically. We conduct a comprehensive analysis of the characteristics of WavCaps dataset and evaluate it on multiple downstream audio-language multimodal learning tasks. The systems trained on WavCaps outperform previous state-of-the-art (SOTA) models by a significant margin. Our aspiration is for the WavCaps dataset we have proposed to facilitate research in audio-language multimodal learning and demonstrate the potential of utilizing ChatGPT to enhance academic research. Our dataset and codes are available at https://github.com/XinhaoMei/WavCaps.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Audio Classification | ESC-50 | Accuracy97.2 | 325 | |
| Text-to-Audio Retrieval | AudioCaps (test) | Recall@142.2 | 145 | |
| Audio Captioning | AudioCaps (test) | CIDEr78.7 | 140 | |
| Audio Classification | Urbansound8K | Accuracy63.6 | 116 | |
| Audio Classification | ESC-50 (test) | Accuracy94.25 | 84 | |
| Audio-to-Text Retrieval | Clotho (test) | R@126.9 | 78 | |
| Musical Instrument Classification | NSynth | Accuracy74.4 | 75 | |
| Audio Classification | SPC V2 | Accuracy73.3 | 65 | |
| Audio-to-Text Retrieval | AudioCaps (test) | R@154.6 | 62 | |
| Text-to-Audio Retrieval | Clotho (test) | R@119.7 | 62 |