Russian scientists created AI system capable of simultaneously analysing speech, facial expressions and voice
Researchers from Russia have developed an artificial intelligence system capable of more accurately extracting information about the user's emotions as it can simultaneously analyse facial expressions, voice and speech, TASS reports
The technology is already showing excellent results in tests, and it can be adapted for virtual assistants, security systems and telemedicine in the future. Its main advantage is flexibility, the audiovisual model works even with insufficient data – i.e. when a face is not visible or a voice is hard to hear.
It is noted that the structure of the system allows it to more accurately account for changes in the emotional state over time. Unlike other emotion recognition systems, the development of Russian scientists is able to take into account several channels of information at once, including data on facial expressions, features in the user's voice and the structure of speech.