Xiaomi’s “Own My Voice” project allows people with speech impairments to have unique and customized voices. knows.
Xiaomi made a sound today with a very nice project. Making a statement on this subject, the company introduced the advanced algorithm application and speech technology it developed for the field of accessibility. Spontaneous style developed by Xiaomi AI Lab Text-to-Speech technology, it is described as producing a unique and customized sound for people with speech disorders. The company cites the following for its important project: “People with speech disorders are able to communicate with others using their ‘own voice’ instead of a monotonous electronic voice. As part of the “Own My Voice” preliminary research project led by the Xiaomi Technical Committee, this successful initiative demonstrates Xiaomi’s commitment to “Technology for Good” and its mission to “Enable everyone around the world to live a better life through innovative technologies” ”
The company explains why they started this project as follows: “Caring for people, Xiaomi tries to meet people’s different needs with technology innovations. Seeing the desire of many people with speech disorders to have unique voices in daily communication, Xiaomi established the “Own My Voice” project team for users to have their own voice. Zhu Xi, Technology for Good Organizer of the Xiaomi Corporation Technology Committee, said, “We are very excited to discover the many values that technology innovations provide us, such as responding to people’s demands.” said.
YOU MAY BE INTERESTED
The company explains how they implemented this project as follows: “Xiaomi worked with more than 200 volunteers within the company to create the most appropriate and personalized voice for the person with speech impairment. A voice trace matching algorithm was used to match the characteristics of the voices donated by the volunteers with the voices of people with speech disorders. With this approach, the most suitable sound was found as the base sound of the voice reference of the person suffering from speech impairment.
Selected with personalization and privacy in mind, the actual sound was manipulated with complex acoustic modification to create a new and original sound. Next, spontaneous-style text-to-speech technology was used to train the AI model, allowing the new voice to gradually acquire a natural rhythm and intonation that could accurately express a person’s emotions and tone.”
According to Xiaomi, the backbone of the project is a group of speech technology experts working within the Xiaomi AI Lab. It is reported that since 2017, these experts have published 37 papers on speech at top international conferences such as the International Conference on Acoustics, Speech and Signal Processing (ICASSP). The success of the “Own My Voice” project is reportedly mainly due to the spontaneous-style Text-to-Speech technology developed by these experts.
YOU MAY BE INTERESTED
In this regard, Xiaomi conveys the following: “Spontaneous Text-to-Speech technology turns the synthesized voice into a real human voice in terms of intonation, pause, velocity and other features. This replaces the monotonous and unnatural feel of electronic sound with a more natural one. Currently, this technology is used in many smart devices equipped with Xiaoai, Xiaomi’s artificial intelligence-supported voice assistant. The “Own My Voice” project demonstrates that spontaneous-style Text-to-Speech technology can be widely adopted in accessibility spaces and improve user experience.”
speaking for the process Zhu Xi, “If we recognize and meet the needs of minority groups at an early stage, the diffusion process of technology can be significantly shortened. This allows the benefits of new technologies to be made available to people with special needs without delay.” says. Xiaomi will continue to receive feedback from people with speech disorders in the future and will examine the feasibility of this project in a wider range.