发明名称 Speech to text training method and system
摘要 An illustrative method includes receiving, at a processor of a computing device, an audio voice signal of a first call participant during a first call, where the first call is a communication across a communication network. The method further includes determining an identity of the first call participant and determining a speech to text profile associated with the identity of the first call participant, where the speech to text profile includes at least one rule for transcribing a word in the audio voice signal into text. The method further includes generating a text output, where the text output is a transcribed version of a plurality of words identified in the audio voice signal of the first call participant. At least one of the plurality of words identified is identified using the at least one rule.
申请公布号 US9444934(B2) 申请公布日期 2016.09.13
申请号 US201414505111 申请日期 2014.10.02
申请人 Nedelco, Inc. 发明人 Nelson Phillip C.;Nelson John;Warren Gerald D.
分类号 H04M3/493;G09B21/00;G10L15/26;H04M3/42;G10L15/22;G10L17/00 主分类号 H04M3/493
代理机构 Foley & Lardner LLP 代理人 Foley & Lardner LLP
主权项 1. A method comprising: receiving, at a processor of a computing device, an audio voice signal of a first call participant during a first call, wherein the first call is a communication across a communication network; determining, by the processor of the computing device, an identity of the first call participant; determining, by the processor of the computing device, a speech to text profile associated with the identity of the first call participant, wherein the speech to text profile comprises a plurality of rules, and further wherein each of the plurality of rules is a rule for transcribing a word in the audio voice signal into text; and wherein the speech to text profile is adequately trained when a number of the plurality of rules reaches a predetermined threshold; and generating, by the processor of the computing device, a text output, wherein the text output is a transcribed version of a plurality of words identified in the audio voice signal of the first call participant, and further wherein at least one of the plurality of words identified is identified using at least one rule of the plurality of rules; and wherein the speech to text profile is used to generate a text output after the speech to text profile has been adequately trained.
地址 Aurora NE US