The Benefit Of Temporally-Strong Labels In Audio Event Classification
This addresses the challenge of label precision in audio datasets for researchers and practitioners, though it is incremental as it builds on existing datasets and methods.
The study tackled the problem of improving audio event classification by using temporally precise ground truth labels, showing that fine-tuning with a mix of weak and strong labels substantially boosts classifier performance, with d' improving from 1.13 to 1.41 on strong evaluation data.
To reveal the importance of temporal precision in ground truth audio event labels, we collected precise (~0.1 sec resolution) "strong" labels for a portion of the AudioSet dataset. We devised a temporally strong evaluation set (including explicit negatives of varying difficulty) and a small strong-labeled training subset of 67k clips (compared to the original dataset's 1.8M clips labeled at 10 sec resolution). We show that fine-tuning with a mix of weak and strongly labeled data can substantially improve classifier performance, even when evaluated using only the original weak labels. For a ResNet50 architecture, d' on the strong evaluation data including explicit negatives improves from 1.13 to 1.41. The new labels are available as an update to AudioSet.