The Center for Artificial Intelligence at MSU has developed a multimodal framework called HL-EAI. According to the university's press service, the platform assesses the emotional state of dialogue participants and formulates responses based on the context. The system operates in three formats: human-AI, human-human, and algorithm-algorithm.
Modern analogues recognize individual emotional signals but do not account for the mutual exchange of feelings in a real conversation. The Russian development combines textual, audio, and visual data. Scientists are already testing the platform in game theory scenarios to measure the impact of emotions on trust and cooperation.
Emotional AI is the next stage after generative AI. Machines that sense the interlocutor and adjust the tone of the response are in demand in negotiations, psychotherapy, and drone control. Western analogues (Affectiva, Empath) dominate the market. The Russian open framework can provide the industry with an independent tool for creating truly lifelike dialogue systems.
