You have a device on the wrist of the hand such as the wearables, can mean several things if there is any company that is able to innovate enough. This wearable is always subject to our body as opposed to a mobile device, which is usually at various times of the day in your pocket or bag.
It is what has come up with more or less the Institute of technology of Massachusetss (MIT) that has shown a demo of a wearable AI system that can predict if a conversation is happy, sad or neutral based on patterns and vital parts of the speech of a person.
Own MIT says:
Matched with vital data and audio, deep learning system could serve in future as a “social coach” for people with anxiety, or Asperger’s syndrome.
Mohammad Ghassemi and Tuka Alhanai graduate student created an algorithm that can analyze speech and tone. These researchers used a 5S iPhone to record the part of the audio to listen to the conversation that allows the system to analyze audio, text transcripts, and physiological signals to determine tone in general history with 83 percent accuracy.
By using the techniques of deep learning / deep learning, the system can also provide a “score of sentiment” in five seconds into the conversation. It was implemented in a Samsung Simband that can measure metrics such as motion, heart rate, blood pressure and skin temperature. The system captured audio data and text transcripts to analyse vocabulary, power and tone of the speaker.
At the moment, the algorithm not can be used for social training, although they are actively working on this objective. Plans of the team is to gather data at a larger scale, so offer in commercial devices of different brands.