Speech recognition system using neural networks
First Claim
1. A speech recognition system comprising:
- voice recognizing and processing means including a plurality of speech recognition neural networks that have previously learned different voice patterns to recognize given voice data, each of said speech recognition neural networks including means for judging whether or not a piece of input voice data coincides with one of the voice data to be recognized and outputting a recognition result and having means for outputting adaptation judgment data independent of the recognition result, the adaptation judgement data representing the adaptation in speech recognition;
selector means receiving input voice data and data from said neural networks and responsive to the adaptation judgment data from each of said speech recognition neural networks for selecting one of said neural networks that has the highest adaptation in speech recognition; and
output control means for outputting the result of speech recognition from the speech recognition neural network selected by said selector means.
0 Assignments
0 Petitions
Accused Products
Abstract
A speech recognition system can recognize a plurality of voice data having different patterns. The speech recognition system has a voice recognizing and processing device including a plurality of speech recognition neural networks that have previously learned different voice patterns to recognize given voice data. Each of the speech recognition neutral networks is adapted to judge whether or not input voice data coincides with one of the voice data to be recognized. Each neural network then outputs adaptation judgment data representing the adaptation in speech recognition. A selector responsive to the adaptation judgment data from each of the speech recognition neural networks selects one of the neural networks that has the highest adaptation in speech recognition. An output control device outputs the result of speech recognition from the speech recognition neural network selected by the selector.
-
Citations
26 Claims
-
1. A speech recognition system comprising:
-
voice recognizing and processing means including a plurality of speech recognition neural networks that have previously learned different voice patterns to recognize given voice data, each of said speech recognition neural networks including means for judging whether or not a piece of input voice data coincides with one of the voice data to be recognized and outputting a recognition result and having means for outputting adaptation judgment data independent of the recognition result, the adaptation judgement data representing the adaptation in speech recognition; selector means receiving input voice data and data from said neural networks and responsive to the adaptation judgment data from each of said speech recognition neural networks for selecting one of said neural networks that has the highest adaptation in speech recognition; and output control means for outputting the result of speech recognition from the speech recognition neural network selected by said selector means. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23)
-
-
24. A speech recognition system comprising:
-
feature extracting means for cutting and convert input voice data into a feature vector for each frame, said feature vectors being sequentially outputted from said feature extracting means; voice recognizing and processing means including a plurality of speech recognition neural networks each having learned to infer a feature vector of a speaker based on a feature vector of a speaker inputted from said feature extracting means into that speech recognition neural network for outputting that inferred vector as adaptation judgement data representing the adaption in the speech recognition, said each speech recognition neural network being formed to output said adaptation judgement data based on a feature vector actually inputted from said feature extracting means; and speaker recognizing means for computing the rate of coincidence between the adaptation judgment data from each of said speech recognition neural network means and the feature vector of the speaker actually inputted from said feature extracting means into said each speech recognition neural network to recognize the speaker of the inputted voice for each of said speech recognition neural network. - View Dependent Claims (25, 26)
-
Specification