Mood_Based_Music_Recommendation_System_with_Facial_Recognition_and_Spotify_API

Utilised facial recognition (based on Deep Neural Network) and similarity based filtering algorithm to recommend songs based on users' current mood.

Techniques Employed

Deep Neural Networks, Facial Recognition, Similarity Based Filtering, Recommendation Algorithm, K-Means Clustering, Euclidian Distance

Tools & Packages Employed

FER (Facial Recognition), Video Slicer, Spotify API, sklearn

Context

This recommendation system will prompt for user to submit a video which will then be processed to determine her/his mood. The user then selects whether he/she prefers to receive a recommendation for songs of similar or opposing moods. Once the mood is determined, the system will then recommend the top 10 songs which are of similar or opposing moods (as selected).

Collaborators

Widya Salim (@salimwid)
Gino Martelli Tiu (@ginosytiu)
Sae Jin Jang (@saejin123)
Susan Koruthu (@skoruthu)