Machine-Learning Aided Device Promises Voice Recovery

Sground-breaking device developed by University of California Los Angeles (UCLA) engineers promises to help individuals with voice disorders speak again.

The team, led by Jun Chen, an assistant professor of bioengineering, has created a soft, thin patch that attaches to the skin outside the throat. The bioelectric system uses machine-learning technology to translate throat muscle movements into speech with nearly 95 per cent accuracy. Detailed in Nature Communications, the device comprises a self-powered sensor that detects muscle movements and an actuator that converts these into audible speech. The innovation, part of Chen’s broader efforts to aid those with disabilities, marks a significant advance in non-invasive voice recovery solutions.

Nvidia launches new chip for AI and Graphic cards

Nvidia unveiled its latest Blackwell platform, is setting a new benchmark in the AI industry with chips designed for real-time generative AI, boasting up to 25x more energy and cost efficiency. Named after mathematician David Harold Blackwell, this technology is poised to dramatically enhance AI’s capabilities in various sectors. The chips for the platform will also find its way into consumer cards later this year. Additionally, Nvidia is venturing into robotics with Project Gr00t.