Wav2Vec2
E435883
automatic speech recognition model
deep learning model
self-supervised learning model
speech representation learning model
Wav2Vec2 is a self-supervised deep learning model for automatic speech recognition that learns powerful audio representations directly from raw waveforms.
All labels observed (7)
| Label | Occurrences |
|---|---|
| Wav2Vec2 canonical | 1 |
| XLSR-53 | 1 |
| wav2vec 2.0 | 1 |
| wav2vec 2.0 Base | 1 |
| wav2vec 2.0 Large | 1 |
| wav2vec 2.0 XLSR | 1 |
| wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations | 1 |
How this entity was disambiguated
This entity first appeared as the object of triple T4389211 — resolving that mention is where its identity was fixed. The disambiguator weighed these candidate entities and picked the highlighted one (or “None”, minting a new entity). This is how homonymy is resolved: the same surface form can point to different entities.
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Wav2Vec2 Context triple: [Hugging Face Transformers, supportsModelType, Wav2Vec2]
-
A.
WaveNet
WaveNet is a deep generative neural network architecture for raw audio that produces highly natural-sounding speech and other audio signals.
-
B.
WaveRNN
WaveRNN is a neural network-based audio waveform generator designed as a more efficient, real-time alternative to earlier autoregressive models for tasks like text-to-speech synthesis.
-
C.
Parallel WaveNet
Parallel WaveNet is a neural vocoder architecture that accelerates high-fidelity audio waveform generation by distilling the autoregressive WaveNet model into a fast, parallelizable form.
-
D.
WaveGlow
WaveGlow is a flow-based generative neural network model for fast, high-quality text-to-speech audio synthesis.
-
E.
GPT-2
GPT-2 is a large transformer-based language model known for generating coherent, human-like text and sparking widespread discussion about the implications of advanced AI text generation.
- F. None of above. chosen
- G. Unsure - the case is ambiguous/there is not enough information to decide.
NED2
Entity disambiguation (via description)
gpt-5-mini-2025-08-07
Target entity: Wav2Vec2 Target entity description: Wav2Vec2 is a self-supervised deep learning model for automatic speech recognition that learns powerful audio representations directly from raw waveforms.
-
A.
WaveNet
WaveNet is a deep generative neural network architecture for raw audio that produces highly natural-sounding speech and other audio signals.
-
B.
WaveRNN
WaveRNN is a neural network-based audio waveform generator designed as a more efficient, real-time alternative to earlier autoregressive models for tasks like text-to-speech synthesis.
-
C.
Parallel WaveNet
Parallel WaveNet is a neural vocoder architecture that accelerates high-fidelity audio waveform generation by distilling the autoregressive WaveNet model into a fast, parallelizable form.
-
D.
WaveGlow
WaveGlow is a flow-based generative neural network model for fast, high-quality text-to-speech audio synthesis.
-
E.
GPT-2
GPT-2 is a large transformer-based language model known for generating coherent, human-like text and sparking widespread discussion about the implications of advanced AI text generation.
- F. None of above. chosen
Statements (48)
| Predicate | Object |
|---|---|
| instanceOf |
automatic speech recognition model
ⓘ
deep learning model ⓘ self-supervised learning model ⓘ speech representation learning model ⓘ |
| achievesStateOfTheArtOn | LibriSpeech 100h setting (at time of publication) ⓘ |
| availableVia | Hugging Face Transformers ⓘ |
| basedOn |
convolutional neural networks
ⓘ
transformer architecture ⓘ |
| developedBy |
Facebook AI Research
ⓘ
Meta AI ⓘ |
| domain |
audio representation learning
ⓘ
speech processing ⓘ |
| fineTuningDataType | labeled speech with transcripts ⓘ |
| hasComponent |
convolutional feature encoder
ⓘ
quantization module ⓘ transformer context network ⓘ |
| hasVariant |
XLSR-53
ⓘ
linked to:
Wav2Vec2
wav2vec 2.0 Base ⓘ
linked to:
Wav2Vec2
wav2vec 2.0 Large ⓘ
linked to:
Wav2Vec2
wav2vec 2.0 XLSR ⓘ
linked to:
Wav2Vec2
|
| implementedIn |
Fairseq
ⓘ
PyTorch ⓘ |
| inputType | 16 kHz mono audio ⓘ |
| inspired |
HuBERT
ⓘ
WavLM ⓘ |
| introducedIn | 2020 ⓘ |
| introducedInPaper |
wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations
ⓘ
linked to:
Wav2Vec2
|
| languageCoverage |
English
ⓘ
multilingual (via XLSR variants) ⓘ |
| learningParadigm |
contrastive learning
ⓘ
self-supervised learning ⓘ |
| operatesOn | raw audio waveforms ⓘ |
| outperforms | previous self-supervised speech models on LibriSpeech ⓘ |
| paperAuthors |
Abdelrahman Mohamed
ⓘ
Alexei Baevski ⓘ Henry Zhou ⓘ Michael Auli ⓘ |
| pretrainingDataType | unlabeled speech audio ⓘ |
| publishedAtConference |
NeurIPS 2020
ⓘ
linked to:
NeurIPS
|
| releasedAs | open-source model ⓘ |
| supportsTask |
keyword spotting
ⓘ
speech classification ⓘ speech recognition fine-tuning ⓘ |
| task | automatic speech recognition ⓘ |
| trainingStrategy | pretrain-then-finetune ⓘ |
| usesMasking | time-step masking on latent speech representations ⓘ |
| usesObjective |
contrastive loss
ⓘ
masked prediction ⓘ |
How these facts were elicited
The pipeline generated the facts above by prompting gpt-5.1 with this entity's name + description and the instruction below.
Instruction
You are a knowledge base construction expert. Given a subject entity and a description of it, return factual statements that you know for the subject as a JSON list of dictionaries(triples), where keys must be "subject", "predicate" and "object". The number of facts may be very high, between 25 to 50 or more, for very popular subjects. For less popular subjects, the number of facts can be very low, like 5 or 10. # Requirements - If you don't know the subject at all, return an empty list. - If the subject is not a named entity, return an empty list. - Include at least one triple where predicate is "instanceOf". - Do not get too wordy. - Separate several objects into multiple triples with one object.
Input
Subject: Wav2Vec2 Description of subject: Wav2Vec2 is a self-supervised deep learning model for automatic speech recognition that learns powerful audio representations directly from raw waveforms.
Referenced by (7)
Full triples — surface form annotated when it differs from this entity's canonical label.
Wav2Vec2
→
introducedInPaper
→
wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations
ⓘ
linked to: Wav2Vec2
linked to: Wav2Vec2
linked to: Wav2Vec2
linked to: Wav2Vec2
linked to: Wav2Vec2