Scholarlink

The rising challenges in mental health are on the rise and need novel approaches. Music therapy aids with mood regulation; however, applying it personally can be complicated. This study presents a solution in form of a system that suggests pre-created songs depending on the user’s emotions for mental health assistance. A CNN-based emotion recognition model processes facial expressions and the recognized emotion is matched with a selected set of songs using Russell’s circumplex model of affect. Each song is accosted with emotional responses and thus the system is able to boost positive moods or reduce negative moods. The system achieves accurate emotion recognition and suitable music choice using the FER2013 dataset, which emphasizes the role of music in different forms of treatment.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Emotion-Based Music Recommendation System

  • 章节

  • 发表日期：

    2026年

查看全文
  • T. Ishwarya,
  • S. S. Nirutthyusha Sri,
  • K. V. Uma,
  • S. Sridevi

摘要

The rising challenges in mental health are on the rise and need novel approaches. Music therapy aids with mood regulation; however, applying it personally can be complicated. This study presents a solution in form of a system that suggests pre-created songs depending on the user’s emotions for mental health assistance. A CNN-based emotion recognition model processes facial expressions and the recognized emotion is matched with a selected set of songs using Russell’s circumplex model of affect. Each song is accosted with emotional responses and thus the system is able to boost positive moods or reduce negative moods. The system achieves accurate emotion recognition and suitable music choice using the FER2013 dataset, which emphasizes the role of music in different forms of treatment.