Decoding multiple HMM sets using a single sentence grammar

• US 7,464,033 B2
• Filed: 02/04/2005
• Issued: 12/09/2008
• Est. Priority Date: 07/31/2000
• Status: Expired due to Term

• Current Assignee: Texas Instruments, Inc.
• Original Assignee: Texas Instruments, Inc.
• Inventors: Gong, Yifan
• Primary Examiner(s): Chawan; Vijay B
• Application Number:
  US11/051,922
• Publication Number:
  US 20050187771A1
• Time in Patent Office: 1,404 Days
• Field of Search: 704255-257, 704231-232, 704/251, 704/240, 704/243, 704/242
• US Class Current: 704/256.2
• CPC Class Codes: G10L 15/142 Hidden Markov Models [HMMs]

Abstract

For a given sentence grammar, speech recognizers are often required to decode M sets of HMMs each of which models a specific acoustic environment. In order to match input acoustic observations to each of the environments, typically recognition search methods require a network of M sub-networks. A new speech recognition search method is described here, which needs a network that is only the size of a single sub-network and yet provides the same recognition performance, thus reducing the memory requirements for network storage by (M-1)/M.

Claims

1. A speech recognizer comprising:

means for receiving speech frames;

means for constructing recognition paths for the speech frames, wherein the recognition paths are defined on expanded-symbols wherein each expanded-symbol references a model contained in one of a plurality of model sets, wherein each model set of the plurality of model sets is a group of models from one acoustic environment;

means for determining expanded-symbols by a conversion function that maps a base-symbol of a generic base grammar network composed of base-symbols to a plurality of expanded-symbols and an expanded-symbol to its corresponding base-symbol; and

means for selecting recognition paths of the constructed recognition paths as containing recognized utterances.

2. The recognizer of claim 1 wherein said recognition path construction includes means for constraining each recognition path to expanded-symbols referencing models within one model set.

3. The recognizer of claim 1 wherein the model sets are HMM models.

4. The recognizer of claim 1 wherein the models of each set correspond to a single environmental factor.

5. The recognizer of claim 1 wherein the recognition procedure consists of a recognition path construction procedure and an update observation probability procedure.

6. The recognition path construction procedure of claim 5 wherein construction of recognition paths consists of extending the path, wherein the path defined by a present expanded-symbol and its referenced model is extended within the referenced model by a within-model-path procedure and to additional expanded-symbols by a cross-model-path procedure.

7. The cross-model-path procedure of claim 6 in which the path from a present expanded-symbol is extended to additional expanded-symbols by determining the present base-symbol corresponding to the present expanded-symbol, determining which additional base-symbols may follow the present-base symbol according to the generic base grammar network, and determining the additional expanded-symbols using the conversion function which maps each additional base-symbol of the generic base grammar network to a plurality of additional expanded-symbols.

8. The update observation probability procedure of claim 5 in which the probability of speech is included in the extended recognition paths for each of the models corresponding to expanded-symbols on the recognition paths.