QuartzNet

GPTKB entity

Statements (32)
Predicate Object
gptkbp:instanceOf speech recognition model
gptkbp:architecture gptkb:convolutional_neural_network
gptkbp:arXivID 1910.10261
gptkbp:author Boris Ginsburg
Jason Kuchaiev
Jing Li
Rene Bolya
Ryan Leary
Samuel Kriman
Stanislav Beliaev
Vitaly Lavrukhin
Yashesh Gaur
gptkbp:availableOn gptkb:NVIDIA_NeMo
gptkbp:basedOn Time-Channel Separable Convolutions
gptkbp:developedBy gptkb:NVIDIA
https://www.w3.org/2000/01/rdf-schema#label QuartzNet
gptkbp:implementedIn gptkb:PyTorch
gptkbp:input audio waveform
gptkbp:introducedIn 2019
gptkbp:language English
gptkbp:notableFor parameter efficiency
high accuracy with fewer parameters
gptkbp:notablePublication QuartzNet: Deep Automatic Speech Recognition with 1D Time-Channel Separable Convolutions
gptkbp:openSource yes
gptkbp:output text transcription
gptkbp:publicationDate gptkb:arXiv
gptkbp:relatedTo gptkb:Jasper
gptkb:Wav2Vec
DeepSpeech
gptkbp:usedFor automatic speech recognition
gptkbp:bfsParent gptkb:NeMo_ASR
gptkbp:bfsLayer 7