Phoneme Score Accelerator

• US 20140180694A1
• Filed: 12/21/2012
• Published: 06/26/2014
• Est. Priority Date: 06/06/2012
• Status: Active Grant

• Current Assignee: Cypress Semiconductor Corporation (Infineon Technologies AG)
• Original Assignee: Spansion LLC (Infineon Technologies AG)
• Inventors: FASTOW, Richard M., Olson, Jens, Bapat, Ojas A.
• Granted Patent: US 9,514,739 B2
• Time in Patent Office: Days
• US Class Current: 704/256.1
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit
• CPC Class Codes: G10L 15/187 Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
• CPC Class Codes: G10L 15/28 Constructional details of speech recognition systems
• CPC Class Codes: G10L 2015/025 Phonemes, fenemes or fenones being the recognition units

Abstract

Embodiments of the present invention include an acoustic processing device and a method for traversing a Hidden Markov Model (HMM). The acoustic processing device can include a senone scoring unit (SSU), a memory device, a HMM module, and an interface module. The SSU is configured to receive feature vectors from an external computing device and to calculate senones. The memory device is configured to store the senone scores and HMM information, where the HMM information includes HMM IDs and HMM state scores. The HMM module is configured to traverse the HMM based on the senone scores and the HMM information. Further, the interface module is configured to transfer one or more HMM scoring requests from the external computing device to the HMM module and to transfer the HMM state scores to the external computing device.

Claims

1. An acoustic processing device for traversing a Hidden Markov Model (HMM), the device comprising:

a senone scoring unit (SSU) configured to receive feature vectors from an external computing device and to calculate senone scores;

a memory device configured to store the senone scores and HMM information, wherein the HMM information comprises HMM IDs and HMM state scores;

an HMM module configured to traverse the HMM based on the senone scores and the HMM information; and

an interface module configured to transfer one or more HMM scoring requests from the external computing device to the HMM module and to transfer the HMM state scores to the external computing device.

2. The acoustic processing device of claim 1, wherein at least one of the HMMs has a variable length.

3. The acoustic processing device of claim 1, wherein the SSU, HMM module, and interface module are integrated on a first chip and the memory device is integrated on a second chip.

4. The acoustic processing device of claim 1, wherein the HMM module is configured to traverse the HMM while the external computing device searches for additional HMMs to evaluate and sends additional HMM scoring requests to the interface module.

5. The acoustic processing device of claim 1, wherein the HMM module is configured to evaluate one or more active lists that include the HMM information.

6. The acoustic processing device of claim 1, wherein the SSU is configured to calculate the senone scores based on an acoustic model.

7. The acoustic processing device of claim 1, wherein the SSU is configured to calculate the senone scores for a frame of data while the external computing device searches for additional HMMs to evaluate and sends additional HMM scoring requests to the interface module.

8. The acoustic processing device of claim 1, wherein the HMM module is configured to evaluate new HMM information associated with a patch list, wherein the new HMM information is different from the HMM information stored in the memory device.

9. The acoustic processing device of claim 1, wherein the SSU is configured to calculate the senone scores prior to the calculation of the HMM state scores by the HMM module.

10. The acoustic processing device of claim 1, wherein the HMM module comprises:

a first buffer configured to store new HMM information, wherein the new HMM information comprises new HMM IDs and new HMM state scores;

a second buffer configured to store the HMM state scores and to transfer the HMM state scores to the external computing device;

a histogram pruning module configured to apply a pruning algorithm to one or more HMM states to generate one or more active HMM states;

an HMM scoring module configured to generate the HMM state scores corresponding to the one or more active HMM states;

a histogram generator configured to generate a distribution of HMM state scores for calculation of a pruning threshold;

a pre-pruning module configured to designate the one or more active HMM states as inactive if the corresponding one or more HMM state scores is below a pre-pruning threshold; and

an arc generator configured to determine whether each of the one or more active HMM states with HMM state scores above the pre-pruning threshold is the last state of an HMM.

11. The acoustic processing device of claim 10, wherein the first buffer, the second buffer, the histogram pruning module, the HMM scoring module, the histogram generator, and the pre-pruning module traverse the HMM in a pipelined manner.

12. A method for traversing a Hidden Markov Model (HMM), the method comprising:

receiving feature vectors from an external computing device;

calculating, with a senone scoring unit (SSU), senone scores;

storing, in a memory device, the senone scores and HMM information, wherein the HMM information comprises HMM IDs and HMM state scores;

applying, with an HMM module, the senone scores and the HMM information when traversing the HMM; and

transferring one or more HMM scoring requests from the external computing device to the HMM module and the HMM state scores to the external computing device.

13. The method of claim 12, wherein the applying comprises traversing the HMM while the external computing device searches for additional HMMs to evaluate and send additional HMM scoring requests.

14. The method of claim 12, wherein the applying comprises evaluating one or more active lists that include the HMM information.

15. The method of claim 12, wherein the calculating comprises calculating the senone scores based on an acoustic model.

16. The method of claim 12, wherein the calculating comprises calculating the senone scores for a frame of data while the external computing device searches for additional HMMs to evaluate and sends additional HMM scoring requests to the interface module.

17. The method of claim 12, wherein the applying comprises evaluating new HMM information associated with a patch list, wherein the new HMM information is different from the HMM information stored in the memory device.

18. The method of claim 12, wherein the calculating comprises calculating the senone scores prior to the calculation of the HMM state scores by the HMM module.

19. The method of claim 12, wherein the applying comprises:

storing new HMM information, wherein the new HMM information comprises new HMM IDs and new HMM state scores;

applying a pruning algorithm to one or more HMM states to generate one or more active HMM states;

generating the HMM state scores corresponding to the one or more active HMM states;

generating a distribution of HMM state scores for calculation of a pruning threshold;

designating the one or more active HMM states as inactive if the corresponding one or more HMM state scores is below a pre-pruning threshold; and

determining whether each of the one or more active HMM states with HMM state scores above the pre-pruning threshold is the last state of an HMM.

20. The method of claim 12, wherein the storing, applying, generating the HMM state scores, generating the distribution of HMM state scores, designating, and determining are performed in a pipelined manner to traverse the HMM.