Statements (25)
Predicate | Object |
---|---|
gptkbp:instanceOf |
speech recognition system
|
gptkbp:application |
automatic speech recognition
|
gptkbp:architecture |
end-to-end neural network
|
gptkbp:basedOn |
deep learning
|
gptkbp:citation |
high
|
gptkbp:developer |
gptkb:Baidu
|
gptkbp:firstReleased |
2014
|
gptkbp:hardware |
GPUs
|
gptkbp:hasResearchCenter |
gptkb:Andrew_Ng
|
https://www.w3.org/2000/01/rdf-schema#label |
Baidu Deep Speech
|
gptkbp:input |
audio waveform
|
gptkbp:language |
gptkb:Mandarin_Chinese
English |
gptkbp:notableAchievement |
state-of-the-art accuracy on speech benchmarks (2014)
|
gptkbp:notableFor |
end-to-end speech recognition without hand-engineered features
|
gptkbp:notablePublication |
Deep Speech: Scaling up end-to-end speech recognition
|
gptkbp:openSource |
no
|
gptkbp:output |
text transcription
|
gptkbp:successor |
Deep Speech 2
|
gptkbp:trainer |
gptkb:Connectionist_Temporal_Classification_(CTC)
large-scale speech datasets |
gptkbp:usedIn |
Baidu virtual assistant
Baidu voice search |
gptkbp:bfsParent |
gptkb:Deep_Speech
|
gptkbp:bfsLayer |
6
|