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

SpeechCommands

Benchmarks

Task NameDataset NameSOTA ResultTrend
Single-class unlearningSpeechCommands v2
FA0.14
10
Diagonal Fisher Information Matrix EstimationSpeechCommands
RelMAE0.17
5
Audio ClassificationSpeechCommands v1 v2 (test)
Accuracy95.3
5
Keyword SpottingSpeechCommands raw (test)
Accuracy (0% Drop)71.66
2
Showing 4 of 4 rows