This publication describes techniques for real-time voice-call language translation performed on a computing device. A Translation Manager of the device may enable real-time language translation of a voice input during a voice call by directing a processor(s) to perform local speech-recognition of the voice input using a machine-learned (ML) model to improve an accuracy of the translation and convert the voice input into text that may be exchanged with another user by real-time text (RTT) over an internet protocol (IP). The processor(s) may translate the text of the voice input using a cloud or a local module to reduce a cost associated with and increase a speed of the translation. Using a text-to-speech (TTS) technique, the user may be provided with the translation of the voice input in a text format and/or an audio format.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Wang, Hui and Ramachandran, Amruth, "Real-Time Voice-Call Language Translation", Technical Disclosure Commons, (January 29, 2021)