Emotion-Aware Speech Processing for ASR and TTS

Boson AI is dedicated to advancing cutting-edge AI research, specializing in large language models for interaction and entertainment. The company focuses on developing innovative AI-driven solutions that enhance human-computer interaction, particularly in speech and language processing. One major challenge Boson AI aims to address through this project is improving AI’s ability to detect subtle emotional cues in speech and generate appropriate responses. Current speech models often struggle with nuanced emotional understanding, leading to unnatural or ineffective interactions. By pretraining a highly intelligent speech model, this project seeks to bridge that gap, enabling more natural and emotionally aware AI interactions. The anticipated benefits for Boson AI include enhancing its AI-powered customer service solutions, leading to more accurate recognition of customer intent and emotions. This advancement could improve user satisfaction, streamline support processes, and provide a competitive edge in AI-driven communication technologies. Additionally, the project may contribute to broader applications in mental health support, virtual assistants, and interactive entertainment, strengthening Boson AI’s position as a leader in AI research.

Faculty Supervisor:

Sushant Sachdeva

Student:

Partner:

Boson AI

Discipline:

Computer science

Sector:

Education; Information and cultural industries

University:

University of Toronto

Program: