Parallel WaveNet
E203077
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.
All labels observed (3)
| Label | Occurrences |
|---|---|
| Parallel WaveNet canonical | 4 |
| Parallel WaveNet: Fast High-Fidelity Speech Synthesis | 1 |
| parallel WaveNet | 1 |
How this entity was disambiguated
This entity first appeared as the object of triple T1793247 — 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: Parallel WaveNet Context triple: [WaveNet, ledTo, Parallel WaveNet]
-
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.
WaveGlow
WaveGlow is a flow-based generative neural network model for fast, high-quality text-to-speech audio synthesis.
-
D.
PixelCNN
PixelCNN is a deep generative model that uses convolutional neural networks with autoregressive masking to model and generate images pixel by pixel.
-
E.
Neural Filters
Neural Filters are Adobe Photoshop’s AI-powered tools that apply advanced, machine-learning-based adjustments and creative effects to images with minimal manual editing.
- 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: Parallel WaveNet Target entity description: 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.
-
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.
WaveGlow
WaveGlow is a flow-based generative neural network model for fast, high-quality text-to-speech audio synthesis.
-
D.
PixelCNN
PixelCNN is a deep generative model that uses convolutional neural networks with autoregressive masking to model and generate images pixel by pixel.
-
E.
Neural Filters
Neural Filters are Adobe Photoshop’s AI-powered tools that apply advanced, machine-learning-based adjustments and creative effects to images with minimal manual editing.
- F. None of above. chosen
Statements (48)
| Predicate | Object |
|---|---|
| instanceOf |
neural vocoder architecture
ⓘ
speech synthesis model ⓘ |
| achieves |
real-time speech synthesis on CPUs
ⓘ
real-time speech synthesis on GPUs ⓘ |
| architectureType | non-autoregressive model ⓘ |
| basedOn | WaveNet ⓘ |
| category |
deep generative model
ⓘ
flow-based generative model ⓘ |
| comparedTo | autoregressive WaveNet ⓘ |
| designedFor | neural text-to-speech ⓘ |
| developedBy | DeepMind ⓘ |
| domain |
speech processing
ⓘ
text-to-speech synthesis ⓘ |
| evaluationMetric | Mean Opinion Score ⓘ |
| framework | TensorFlow ⓘ |
| generationMode | parallel sampling ⓘ |
| handles |
16 kHz audio
ⓘ
24 kHz audio ⓘ |
| improvesUpon | WaveNet inference speed ⓘ |
| inputType |
acoustic features
ⓘ
linguistic features ⓘ |
| language | Python ⓘ |
| openSourcedAs | WaveNet implementation variants by DeepMind ⓘ |
| organization | Google ⓘ |
| outputType | raw audio waveform ⓘ |
| property |
fast inference
ⓘ
high-fidelity audio generation ⓘ high-quality speech ⓘ near-human naturalness in MOS evaluations ⓘ parallel waveform generation ⓘ |
| publicationType | research paper ⓘ |
| publicationVenue |
ICML 2018
ⓘ
linked to:
ICML
|
| reduces | generation latency ⓘ |
| relatedTo |
ClariNet
ⓘ
Tacotron ⓘ WaveGlow ⓘ WaveRNN ⓘ |
| studentModel |
parallel WaveNet
ⓘ
linked to:
Parallel WaveNet
|
| task |
speech synthesis
ⓘ
waveform generation ⓘ |
| teacherModel | autoregressive WaveNet ⓘ |
| title |
Parallel WaveNet: Fast High-Fidelity Speech Synthesis
ⓘ
linked to:
Parallel WaveNet
|
| trainingObjective | match teacher WaveNet distribution ⓘ |
| trainingStrategy | teacher-student distillation ⓘ |
| uses |
inverse autoregressive flow
ⓘ
knowledge distillation ⓘ normalizing flows ⓘ probability density distillation ⓘ |
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: Parallel WaveNet Description of subject: 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.
Referenced by (6)
Full triples — surface form annotated when it differs from this entity's canonical label.
linked to: Parallel WaveNet
linked to: Parallel WaveNet