HuBERT
E435884
HuBERT is a self-supervised speech representation learning model that learns powerful audio features from unlabeled speech for tasks like automatic speech recognition and audio classification.
All labels observed (1)
| Label | Occurrences |
|---|---|
| HuBERT canonical | 1 |
How this entity was disambiguated
This entity first appeared as the object of triple T4389212 — 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: HuBERT Context triple: [Hugging Face Transformers, supportsModelType, HuBERT]
-
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.
Hugging Face Transformers
Hugging Face Transformers is a widely used open-source library that provides state-of-the-art transformer-based models and tools for natural language processing and related machine learning tasks.
-
E.
WaveGlow
WaveGlow is a flow-based generative neural network model for fast, high-quality text-to-speech audio synthesis.
- 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: HuBERT Target entity description: HuBERT is a self-supervised speech representation learning model that learns powerful audio features from unlabeled speech for tasks like automatic speech recognition and audio classification.
-
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.
Hugging Face Transformers
Hugging Face Transformers is a widely used open-source library that provides state-of-the-art transformer-based models and tools for natural language processing and related machine learning tasks.
-
E.
WaveGlow
WaveGlow is a flow-based generative neural network model for fast, high-quality text-to-speech audio synthesis.
- F. None of above. chosen
Statements (51)
| Predicate | Object |
|---|---|
| instanceOf |
deep learning model
ⓘ
self-supervised speech representation learning model ⓘ speech foundation model ⓘ |
| basedOn | Transformer architecture ⓘ |
| designedFor |
audio classification
ⓘ
automatic speech recognition ⓘ downstream speech tasks ⓘ self-supervised learning from speech ⓘ speech representation learning ⓘ spoken language understanding ⓘ |
| developedBy |
Facebook AI Research
NERFINISHED
ⓘ
Meta AI NERFINISHED ⓘ |
| evaluationBenchmark |
Libri-light
NERFINISHED
ⓘ
LibriSpeech NERFINISHED ⓘ TIMIT NERFINISHED ⓘ |
| hasAuthor |
Benjamin Bolte
NERFINISHED
ⓘ
James Glass NERFINISHED ⓘ Kyunghyun Cho NERFINISHED ⓘ Wei-Ning Hsu NERFINISHED ⓘ Yao-Hung Hubert Tsai NERFINISHED ⓘ |
| hasVariant |
HuBERT Base
NERFINISHED
ⓘ
HuBERT Large NERFINISHED ⓘ HuBERT X-Large NERFINISHED ⓘ |
| improvesOver | wav2vec 2.0 on several speech benchmarks ⓘ |
| inputType |
acoustic features such as log-mel filterbanks
ⓘ
raw audio waveforms ⓘ |
| introducedIn | 2021 ⓘ |
| introducedInPaper | HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units NERFINISHED ⓘ |
| language | primarily English in original experiments ⓘ |
| learningSignalSource | discrete units obtained by clustering MFCC or filterbank features ⓘ |
| maskingStrategy | masking of contiguous time spans in the input sequence ⓘ |
| openSourceImplementation |
Hugging Face Transformers
NERFINISHED
ⓘ
fairseq NERFINISHED ⓘ |
| outputType |
contextualized speech representations
ⓘ
frame-level audio embeddings ⓘ |
| pretrainingStage |
masked region prediction of cluster assignments
ⓘ
offline clustering of acoustic features ⓘ |
| publishedAt | Interspeech 2021 NERFINISHED ⓘ |
| relatedTo |
Data2Vec
NERFINISHED
ⓘ
WavLM NERFINISHED ⓘ wav2vec 2.0 NERFINISHED ⓘ |
| supportsTask |
audio event classification
ⓘ
automatic speech recognition fine-tuning ⓘ emotion recognition from speech ⓘ keyword spotting ⓘ phoneme recognition ⓘ speaker recognition ⓘ |
| trainingDataType | unlabeled speech audio ⓘ |
| trainingParadigm | self-supervised learning ⓘ |
| usesObjective |
cluster-based prediction task
ⓘ
masked prediction of latent speech units ⓘ |
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: HuBERT Description of subject: HuBERT is a self-supervised speech representation learning model that learns powerful audio features from unlabeled speech for tasks like automatic speech recognition and audio classification.
Referenced by (1)
Full triples — surface form annotated when it differs from this entity's canonical label.