Baevski and colleagues: wav2vec 2.0 self-supervised speech learning

What changed

Learned speech representations from unlabeled audio with masked latent prediction and contrastive training, then fine-tuned with limited transcriptions.

Scientific scope

Learned speech representations from unlabeled audio with masked latent prediction and contrastive training, then fine-tuned with limited transcriptions. Scope limitation: The reported low-label recognition results are on the stated speech datasets; this is not a universal language-independent speech system.

Evidence

The retained primary-source abstract, bibliographic record and stated review scope support this provisional assessment. No independent experimental replication by WRS is claimed. The reported low-label recognition results are on the stated speech datasets; this is not a universal language-independent speech system.

Contributors

• The named authors listed in the linked research record · SCIENCE: Complete source byline retained in the research record; human research direction, method development and evaluation. Individual labor shares are not measured.

• Baevski and colleagues: wav2vec 2.0 self-supervised speech learning — original paper
• Baevski and colleagues: wav2vec 2.0 self-supervised speech learning — official company source

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction