StanislavKo28/music_moods_classification

Music mood classification is a fundamental and versatile application in many various domains. Some possible use cases for music mood classification include:

• music recommendation systems;
• content organization and discovery;
• radio broadcasting and programming;
• music licensing and copyright management;
• music analysis and research;
• content tagging and metadata enrichment;
• audio identification and copyright protection;
• music production and creativity;
• healthcare and therapy;
• entertainment and gaming.

The model is trained based on publicly available dataset of labeled music data — HWNAS Dataset — that contains 6930 sample 30-second audio files evenly split among 14 moods:

• angry;
• dark;
• energetic;
• epic;
• euphoric;
• happy;
• mysterious;
• relaxing;
• romantic;
• sad;
• scary;
• glamorous;
• uplifting;
• sentimental.

Kaggle notebooks:

• Training - Kaggle notebook;
• Inference - Kaggle notebook.