It has been awarded patent number 7,437,290 in which an input audio data stream comprising speech is processed by an automatic censoring filter in either a real-time mode, or a batch mode.
The big idea is that the speech is censored so that undesired words or phrases are either unintelligible or inaudible. It means that if your call centre employee starts swearing at a customer, the punter would never know. If the software was installed on a company phone then it would be impossible to say anything that the company didn’t want you to say.
According to the patent the automatic censoring filter employs a lattice comprising either phonemes and words derived from them. It compares these against a list of words included in undesired speech data.
If there is a match the input audio data stream is altered so that the undesired word or a phrase comprising a plurality of such words is unintelligible or inaudible.
The censored speech can either be stored or made available to an audience in real-time.
Microsoft patents speech censoring
By Nick Farrell on Oct 21, 2008 6:35AM