发明名称 |
METHOD AND APPARATUS FOR SPEECH RECOGNITION |
摘要 |
A computer-implemented method, apparatus and computer program product. The computer-implemented method performed by a computerized device, comprising: transforming a hidden Markov model to qubits; transforming data into groups of qubits, the data being determined upon the hidden Markov model and features extracted from an audio signal, the data representing a likelihood observation matrix representing likelihood of phoneme and state combinations in an audio signal; applying a quantum search algorithm for finding a maximal value of the qubits; and transforming the maximal value of the qubits into a number, the number representing an entry in a delta array used in speech recognition. |
申请公布号 |
US2014222431(A1) |
申请公布日期 |
2014.08.07 |
申请号 |
US201314237128 |
申请日期 |
2013.07.07 |
申请人 |
NOVOSPEECH LTD. |
发明人 |
Ben-Ezra Yossef |
分类号 |
G10L15/14 |
主分类号 |
G10L15/14 |
代理机构 |
|
代理人 |
|
主权项 |
1. A computer-implemented method performed by a computerized device, comprising:
transforming a hidden Markov model to qubits; transforming data into groups of qubits, the data being determined upon the hidden Markov model and features extracted from an audio signal, the data representing a likelihood observation matrix representing likelihood of phoneme and state combinations in an audio signal; applying a quantum search algorithm for fording a maximal value of the qubits; and transforming the maximal value of the qubits into a number, the number representing an entry in a delta array used in speech recognition. |
地址 |
Petach Tikva IL |