Google have created artificial intelligence that is able to distinguish the speech of people, even when they kill each other. In the future it can be integrated into the company’s products.
The developers at Google have created artificial intelligence that can identify a specific voice, looking at people when they talk. The team first taught the AI to recognize the voices separately, and then created the virtual background noise to the AI learned to distin guish each of the voices in different audio tracks. Based on this AI creates one audio track, which features the speech of the speaker. The system works even if the face is covered with your hand or a microphone.
Google explains how you can integrate the development into its own products. AI can be used in Hangouts and the Google Duo to improve sound quality when talking in a crowded room. It can also be useful when recording video. Google believe that AI can be used in hearing AIDS, especially in situations where telling a few people.
Follow the news in our Telegram channel (while you can), as well as in the app on iOS MacDigger.
