AI Music Recommendation System Based on User Emotions and Context

Explore a 14-minute conference talk from SK TECH SUMMIT 2023 that delves into developing an AI-powered music recommendation system that matches users' moods and situations. Learn how FLO developed a multi-modal model that maps text meaning to audio characteristics, enabling personalized playlist recommendations based on users' emotional states and contextual situations. Discover the innovative two-stage recommendation system that combines natural language processing with audio-audio matching to enhance playlist quality. Follow along as speakers Seok-woo Kang and Si-jeong Hwang from Dreamus Company detail the dataset creation, deep learning models, and system architecture behind this emotion-based music recommendation platform. Gain insights into how this technology aims to help users discover music that resonates with their current emotional state while promoting musical diversity beyond trending tracks.