Implementación de un sistema de gestos y voz emocionalmente sincronizados en robots sociales

Abstract

DOCUMENT IN THE PROCESS OF INTELLECTUAL PROPERTY. The project develops a gesture and emotion synchronization system for a robotic bust designed for storytelling aimed at children. This system seeks to overcome current limitations in human-robot interaction by enhancing the robot's ability to capture attention and convey emotions in a natural and effective way. A modular design was employed, including an emotional analysis module based on natural language processing (NLP), voice generation using Azure TTS with SSML for prosody adjustments, and predefined gesture libraries to synchronize movements with the speech. The tests were conducted with 10 eighth-grade students, evaluating attention capture and synchronization perception. The results showed that 100% of participants enjoyed the storytelling, 80% maintained constant visual contact with the robot, and 90% perceived the gestures as coherent with the narrated emotions. However, 50% indicated that the synchronization of the beak with the audio was partial. It is concluded that the system achieves immersive and effective interaction, though there are opportunities to improve lip sync precision and the calibration of certain gestures. This work lays a foundation for expanding the capabilities of social robots in educational and entertainment contexts. Keywords: Human-Robot Interaction, Narrative, Emotional Technology, Expressiveness