PaSST–RoBERTa & GPT-augment

Reported on 4 benchmarks across 1 task · 1 paper · 3 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Audio4 results

Text to Audio RetrievalonClotho
R@10· uses extra data· 2023-08-08
69.3
best: 70.39 (PaSST-RoBERTa & Estimated Audio–Caption Correspondences)
SOTA
Advancing Natural-Language Based Audio Retrieval with PaSST and Large Audio-Caption Data Sets arXiv:2308.04258
Text to Audio RetrievalonClotho
R@5· uses extra data· 2023-08-08
55.27
best: 57.03 (PaSST-RoBERTa & Estimated Audio–Caption Correspondences)
SOTA
Advancing Natural-Language Based Audio Retrieval with PaSST and Large Audio-Caption Data Sets arXiv:2308.04258
Text to Audio RetrievalonClotho
mAP@10· uses extra data· 2023-08-08
38.56
best: 40.14 (PaSST-RoBERTa & Estimated Audio–Caption Correspondences)
SOTA
Advancing Natural-Language Based Audio Retrieval with PaSST and Large Audio-Caption Data Sets arXiv:2308.04258
Text to Audio RetrievalonClotho
R@1· uses extra data· 2023-08-08
26.07
best: 27.69 (PaSST-RoBERTa & Estimated Audio–Caption Correspondences)
Advancing Natural-Language Based Audio Retrieval with PaSST and Large Audio-Caption Data Sets arXiv:2308.04258