18077934. TECHNIQUES FOR SECURELY SYNTHESIZING SPEECH WITH THE NATURAL VOICE OF A SPEAKER DURING A LANGUAGE-TRANSLATED COMMUNICATION SESSION simplified abstract (Microsoft Technology Licensing, LLC)

From WikiPatents
Jump to navigation Jump to search

TECHNIQUES FOR SECURELY SYNTHESIZING SPEECH WITH THE NATURAL VOICE OF A SPEAKER DURING A LANGUAGE-TRANSLATED COMMUNICATION SESSION

Organization Name

Microsoft Technology Licensing, LLC

Inventor(s)

Jan Pavlovsky of Redmond WA (US)

Adam Czeisler of Redmond WA (US)

Luis Carrasco of Seattle WA (US)

TECHNIQUES FOR SECURELY SYNTHESIZING SPEECH WITH THE NATURAL VOICE OF A SPEAKER DURING A LANGUAGE-TRANSLATED COMMUNICATION SESSION - A simplified explanation of the abstract

This abstract first appeared for US patent application 18077934 titled 'TECHNIQUES FOR SECURELY SYNTHESIZING SPEECH WITH THE NATURAL VOICE OF A SPEAKER DURING A LANGUAGE-TRANSLATED COMMUNICATION SESSION

Simplified Explanation: The patent application describes a technique for securely synthesizing a voice of a speaker during a language translated voice call.

  • **Key Features and Innovation:**
   * Audio data representing the speech of the speaker is received and processed at a server computer.
   * A fixed-duration sample of the speaker's speech is obtained and used to generate a voice profile.
   * The voice profile is continuously updated at fixed intervals during the voice call.

Potential Applications: This technology could be used in secure language translation services, virtual meetings, customer service calls, and other communication platforms where voice synthesis is required.

Problems Solved: This technology addresses the need for secure voice synthesis during language translated voice calls, ensuring accurate representation of the speaker's voice in a second language.

Benefits: The benefits of this technology include enhanced communication in multilingual settings, improved accuracy of voice synthesis, and increased security in voice calls.

Commercial Applications: The technology could be applied in language translation services, telecommunication companies, virtual meeting platforms, and customer service call centers to enhance communication and security.

Prior Art: Prior research in voice synthesis, language translation, and secure communication technologies may be relevant to this innovation.

Frequently Updated Research: Stay informed about advancements in voice synthesis, language translation, and secure communication technologies to enhance the capabilities of this innovation.

Questions about Voice Synthesis in Language Translated Voice Calls: 1. How does the continuous updating of the voice profile improve the accuracy of voice synthesis during a call? 2. What measures are in place to ensure the security of the voice data processed during the call?


Original Abstract Submitted

Described herein is a technique for securely synthesizing a voice of a speaker during a language translated voice call. When a voice call is first initiated between the speaker and one or more other call participants, the audio data representing the speech of the speaker is received at a server computer where it is processed by obtaining a sample of a fixed duration (e.g., 8 seconds). This fixed-duration sample is then processed to generate a voice profile of the speaker for use in generating synthesized speech in a voice of the speaker, in a second language. This process of sampling the audio data and generating the voice profile is repeated at a fixed interval (e.g., every 30 seconds), such that the voice profile of the speaker is continuously updated during the voice call.

Retrieved from "http://wikipatents.org/index.php?title=18077934._TECHNIQUES_FOR_SECURELY_SYNTHESIZING_SPEECH_WITH_THE_NATURAL_VOICE_OF_A_SPEAKER_DURING_A_LANGUAGE-TRANSLATED_COMMUNICATION_SESSION_simplified_abstract_(Microsoft_Technology_Licensing,_LLC)&oldid=74028"