List your company

ResearchFormerly WizWand

AudioCaps

Benchmarks

TaskDatasetSOTA resultTrendResults

Audio Captioning

AudioCaps (test)

91.1CIDEr

222

Text-to-audio generation

AudioCaps (test)

0KL Divergence

213

text-to-audio retrieval

AudioCaps (test)

66.59Recall@1

191

Audio-to-text Retrieval

AudioCaps (test)

65.6R@1

80

Audio Captioning

AudioCaps

80.3CIDEr

66

Text-to-audio retrieval

AudioCaps

55.2Recall@1

57

Audio Retrieval

AudioCaps

52R@1

56

AudioCaps

1.86FD (OpenL3)

27

AudioCaps

12.38FD (Frechet Distance)

24

Cross-modal retrieval

AudioCaps (test)

59.1R@1

23

Audio-to-Text Retrieval

AudioCaps

45.1R@1

22

Automated Audio Captioning

AudioCaps (evaluation)

51.8SPIDEr

17

Zero-shot Retrieval (T+V → A)

AudioCaps

95.2Recall@1

14

Zero-shot Retrieval (T+A → V)

AudioCaps

89Recall@1

14

Zero-shot Retrieval (T → A+V)

AudioCaps

45.8Recall@1

14

Text-to-text retrieval

AudioCaps

50.3Recall@1

13

AudioCaps

4.43R-MOS

12

Audio Question Answering

AudioCaps-QA (test)

60.77Model-as-Judge Score

12

environment-aware text-to-speech

AudioCaps (test)

6.76WER

11

Audio Question Answering

AudioCaps (test)

60.1Token-Level Accuracy

11

Audio Understanding

AudioCaps

42.82LB Score

11

Text-to-audio generation

AudioCaps (evaluation)

1.85FAD

11

AudioCaps 2019 (test)

1.558FAD

10

Video-to-Audio Retrieval

AudioCaps V→A

88.3Recall@1

10

Text-to-Video Retrieval

AudioCaps T→V

20.8Recall@1

10