Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

AudioMarathon

Benchmarks

Task NameDataset NameSOTA ResultTrend
Long-Form Speech UnderstandingAudioMarathon 1.0 (test)
Average Score65.6
16
Audio ClassificationAudioMarathon 1.0 (test)
SED Score78.4
15
Speaker Information ModelingAudioMarathon 1.0 (test)
SD (Score)72.3
15
Speech Content ExtractionAudioMarathon 1.0 (test)
SER12.4
15
Sound Event DetectionAudioMarathon
SED F178.4
11
Showing 5 of 5 rows