Search Here

A wearable AI that can predict whether the tone of the conversation is happy, sad or neutral can soon be realized to help fight social anxiety. The system will base its reading on the user's speech patterns and vital signs.

Researchers from Computer Science and Artificial Intelligence Laboratory (CSAIL) of MIT and Insitute of Medical Engineering and Science (IMES) are a step closer in helping people with anxiety disorder or Asperger's condition. Tuka Alhanai, a graduate student and co-author of the study said that the wearable AI will be like a social coach in the user's pocket.

Mohammad Ghassemi, together with Alhanai, will present the new technology in the Association for the Advancement of Artificial Intelligence (AAAI). The wearable AI can analyse audio, text transcriptions and physiological signals. The accuracy of the gadget is around 83 percent. With deep-learning techniques, it can also provide a sentiment score for specific intervals within a conversation for five seconds.

Ghassemi said that the wearable AI can classify the emotional tone of conversations in real-time. While the subjects are having natural and unstructured interaction, it can collect physical and speech data in a passive but robust way.

The performance of the wearable AI will further be enhanced if there are more than two people involved in the conversation wearing the same technology. It creates more data that can be analysed by the algorithms.

Alhanai assured the public that the wearable AI was developed with strong considerations on the user's privacy. The algorithm runs only locally so it can protect personal information.

In the research, the subjects wore a Samsung Simband to capture high-resolution physiological waveforms such as blood flow, blood pressure, heart rate, movement and skin temperature. The wearable AI can also capture audio data and text transcripts to analyse the user's energy, pitch, tone and vocabulary.