AI Paper Atlas

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

DinoSR: Self-Distillation and Online Clustering for Self-supervised Speech Representation Learning

Alexander H. Liu, Heng-Jui Chang, Michael Auli, Wei-Ning Hsu, James R. Glass

NeurIPS 2023 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Self-supervised speech representations · 6 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

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
online clustering with codebook instead of continuous self-distillation (data2vec)DinoSR · ZeroSpeech 2021 ABX averageABX error rate (%): 5.20 -> 3.59Improved
online clustering instead of offline iterative clustering (HuBERT)DinoSR · ZeroSpeech 2021 ABX averageABX error rate (%): 4.22 -> 3.59Improved
DinoSR vs data2vec fine-tuningDinoSR · LibriSpeech test-other, 10 minWER (%): 12.3 -> 11.8Improved
DinoSR vs data2vec fine-tuningDinoSR · LibriSpeech test-other, 100 hrWER (%): 6.8 -> 6.7No clear effect
DinoSR vs data2vec fine-tuningDinoSR · LibriSpeech dev-clean, 100 hrWER (%): 2.2 -> 2.3Hurt
DinoSR vs data2vec on SUPERB PRDinoSR · SUPERB PRPER (%): 4.69 -> 3.21Improved
DinoSR vs WavLM on SUPERB slot filling CERDinoSR · SUPERB SFCER (%): 22.86 -> 23.57Hurt
vary codebook size V from 64 to 2048 (ablation)DinoSR · LibriSpeech dev-other, 10 hr fine-tuneWERNo clear effect
vary the top N layers that apply online clustering (ablation)DinoSR · LibriSpeech dev-other, 10 hr fine-tuneWERMixed
vary codebook decay rate tau between 0.5 and 0.99 (ablation)DinoSR · LibriSpeech dev-other, 10 hr fine-tuneWERNo clear effect
anneal codebook decay tau to 0.99 or 0.999 (ablation)DinoSR · LibriSpeech dev-other, 10 hr fine-tuneWERNo clear effect
DinoSR online codebook vs VQ-APC code usageDinoSR · LibriSpeech dev unit qualitycode perplexity: 72.1 -> 179.2Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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