In a groundbreaking development, researchers at Ulsan National Institute of Science and Technology (UNIST) have created a real-time human emotion recognition technology that has the potential to revolutionize various industries, particularly wearable systems. The team overcame the challenge of processing abstract emotional data by developing a multi-modal system that combines verbal and non-verbal expression data.
The system utilizes a personalized skin-integrated facial interface (PSiFI) powered by friction charging. A bidirectional triboelectric strain and vibration sensor enables simultaneous data sensing and integration, with a fully integrated data processing circuit ensuring wireless real-time emotion recognition, even when individuals are wearing masks.
Through machine learning algorithms, the technology has demonstrated impressive accuracy in recognizing human emotions. This was seen in a digital concierge application within virtual reality (VR) environments. By incorporating data from facial muscle deformation and vocal cord vibrations, the system can provide personalized services tailored to users’ emotions.
This advancement in human-machine interface (HMI) devices opens up possibilities for portable emotion recognition devices and next-generation digital platform services. Collaboration with Nanyang Technical University in Singapore, along with support from the National Research Foundation of Korea (NRF) and the Korea Institute of Materials (KIMS), highlights the significance of this development in HMI devices and sets the stage for improved interaction capabilities between humans and machines.
