AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

A Sequential Self Teaching Approach for Improving Generalization in Sound Event Recognition

Anurag Kumar, Vamsi Ithapu

ICML 2020 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Self-supervised speech and brain decoding · 20 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use sequential co-supervision from previous modelsWEANET · AudioSetmAP: 0.366 -> 0.398Improved
Use two previous-stage teachers instead of one (ablation)SUSTAIN on WEANET · AudioSetmAP: 0.394 -> 0.398Improved
Combine ground-truth labels with teacher predictions rather than using either alone (ablation)SUSTAIN on WEANET · AudioSetmAP: up to 5.7%Improved
Increase the number of sequential stages on the noisy training set (ablation)SUSTAIN with WEANET^L · FSDKaggle Noisylwlrap: 0.436 -> 0.472Improved
Continue co-supervision beyond one stage on clean curated data (ablation)SUSTAIN with WEANET^L · FSDKaggle CuratedlwlrapHurt
Use class-specific attention instead of average pooling (ablation)WEANET · AudioSetmAP: 0.352 -> 0.366Improved
Transfer features from a SUSTAIN-trained model rather than the base modelAudioSet-trained WEANET · ESC-50 transferaccuracy: 92.6 -> 94.1Improved
Transfer features from a SUSTAIN-trained model rather than the base modelAudioSet-trained WEANET · FSDKaggle Noisy transferlwlrap: 0.486 -> 0.503Improved
Transfer features from a SUSTAIN-trained model rather than the base modelAudioSet-trained WEANET · FSDKaggle Curated transferlwlrap: 0.712 -> 0.728Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy