01-03-2017, 11:57 AM
Voice morphing (also known as voice conversion and voice conversion) is the disturbance generated by a person's natural voice software. The purpose may be to add audio effects to the voice, hide the identity of the person or impersonate another person.Morphing voice software that alters a voice to sound like a target guy is the most sophisticated technology and the most difficult to create. The CUED project at Cambridge University explains:
There are basically three interdependent problems that must be resolved before constructing a voice morphing system. In the first place, it is important to develop a mathematical model to represent the voice signal so that synthetic speech can be regenerated and prosody can be manipulated without artifacts. Secondly, the various acoustic signals that allow humans to identify the loudspeakers must be identified and extracted. Third, the type of conversion function and the method of formation and application of the conversion function must be decided.
At the Los Alamos National Laboratory in New Mexico, researchers developed a voice morphing technology that allowed them to create compelling representations of the voices of several US generals. In 1999, Los Alamos researchers replicated the speech patterns of a 10-minute recording of General Carl W. Steiner and a message to the general: "Gentlemen, we have called you to inform you that we are going to overthrow the government of The United States. "Other voice clones included Colin Powell saying" I'm being treated well by my captors. "
There are several voice morphing applications available to consumers. An iPhone application called Voice Morphing offers modes that include helium, gangster, robot and alien. Second Life's morphing voice effects include techie, male, female, scary and minuscule.
There are basically three interdependent problems that must be resolved before constructing a voice morphing system. In the first place, it is important to develop a mathematical model to represent the voice signal so that synthetic speech can be regenerated and prosody can be manipulated without artifacts. Secondly, the various acoustic signals that allow humans to identify the loudspeakers must be identified and extracted. Third, the type of conversion function and the method of formation and application of the conversion function must be decided.
At the Los Alamos National Laboratory in New Mexico, researchers developed a voice morphing technology that allowed them to create compelling representations of the voices of several US generals. In 1999, Los Alamos researchers replicated the speech patterns of a 10-minute recording of General Carl W. Steiner and a message to the general: "Gentlemen, we have called you to inform you that we are going to overthrow the government of The United States. "Other voice clones included Colin Powell saying" I'm being treated well by my captors. "
There are several voice morphing applications available to consumers. An iPhone application called Voice Morphing offers modes that include helium, gangster, robot and alien. Second Life's morphing voice effects include techie, male, female, scary and minuscule.