Apparatus and method for recognizing speech based on a deep-neural-network (DNN) sound model
AssigneeElectronics and Telecommunications Research Institute
InventorsByung Ok Kang, Jeon Gue Park, Hwa Jeon Song, Yun Keun Lee, Eui Sok Chung
Life of this patent
Estimating the term…
Abstract
A speech recognition apparatus based on a deep-neural-network (DNN) sound model includes a memory and a processor. As the processor executes a program stored in the memory, the processor generates sound-model state sets corresponding to a plurality of pieces of set training speech data included in multi-set training speech data, generates a multi-set state cluster from the sound-model state sets, and sets the multi-set training speech data as an input node and the multi-set state cluster as output nodes so as to learn a DNN structured parameter.
Claims16
A speech recognition apparatus based on a deep-neural-network (DNN) sound model, comprising:
- a memory;
- and
- a processor configured to execute a program stored in the memory,
- wherein, as the program is executed, the processor:
- generates sound-model state sets corresponding to a plurality of pieces of set training speech data included in multi-set training speech data,
- generates a multi-set state cluster from the sound-model state sets, generating the multi-set state cluster comprising:
- collecting the sound model states;
- calculating respective state log likelihoods of the sound models states;
- and
- generating the multi-set state cluster by merging a first sound model state set with a second model state set using the state log likelihoods,
- learns a DNN structured parameter by setting the multi-set training speech data as an input node of the DNN, setting the multi-set state cluster as an output node of the DNN, and disconnecting the output node from a state cluster of the DNN not related to the input node, and
- when a user's speech and characteristic information thereof are received via a user interface, recognizes the user's speech on the basis of the learned DNN structured parameter by setting a sound-model state set corresponding to the characteristic information of the user's speech as an output node.
A speech recognition method based on a deep-neural-network (DNN) sound model, comprising:
- generating sound-model state sets corresponding to a plurality of pieces of set training speech data included in multi-set training speech data;
- generating a multi-set state cluster from the sound-model state sets, generating a multi-set state cluster including collecting the sound-model state sets, calculating state log likelihoods of each state of the sound-model state sets, and generating the multi-set state cluster by merging first and second sound-model state sets on the basis of the state log likelihoods and state tying information of the sound-model state sets;
- learning a DNN structured parameter by setting the multi-set training speech data as an input node, setting the multi-set state cluster as an output node, and disconnecting the output node from a state cluster of the DNN not related to the input node;
- receiving a user's speech and characteristic information thereof via a user interface;
- and
- recognizing the user's speech on the basis of the learned DNN structured parameter by setting a sound-model state set corresponding to the characteristic information of the user's speech as an output node.
Dependencies are inferred from the claim text.