Music Moody - Facial Recognition And Voice Recognition To Detect Mood And Recommend Songs
Volume 6 - Issue 10, October 2023 Edition
[Download Full Paper]
Author(s)
K. Tharmikan, Heisapirashoban.N, M.A. Miqdad Ali Riza, R.R. Stelin Dinoshan, Thusithanjana Thilakarthna
Keywords
Collaborative Filtering, Music Recommendations, Streaming services, Digital music libraries, Music preferences, Content-based filter algorithms, User-specific, Data preprocessing, Feature engineering, Model training, Matrix factorization algorithm, Latent features, Playlist creation, Mood Detection, Live Voice Recognition, Speech Emotion Recognition, Neural Network, Machine Learning Algorithms, Vocal Features, Emotional State, Song Recommendations, Audio Analysis, Emotional Resonance, Music Recommendation, Facial Emotion Recognition, Personalized Music, Emotional Context, Mood Detection, Emotion-based Recommendations, Stress Reduction, User Engagement, Data-driven Recommendations, Emotional Resonance, Technology Advancements, User Data.
Abstract
This project aims to develop a comprehensive music recommendation system that provides personalized song suggestions based on the user's individual tastes and current emotional state. The system incorporates four main components: mood detection using live voice recognition techniques, collaborative filtering for playlist generation, multiclassification of songs based on mood, and base and frequency feature extraction. The real-time voice recognition module analyzes the user's voice to extract features like pitch, volume, and tone, which are then used to determine the user's mood state. This information is fed into the mood detection and song recommendation module, which employs a neural network trained on a large dataset of labeled audio recordings to predict the user's mood. Also utilizes collaborative filtering techniques, considering the user's music preferences, listening history, and similarities with other users, to generate personalized song playlists. Additionally, a multiclassification approach using base and frequency features is employed to classify songs into mood categories such as happy, sad, calm, and energetic. This classification allows for better organization and recommendation of songs based on their emotional characteristics. Overall, this project offers a comprehensive approach to personalized music recommendation, leveraging voice recognition, collaborative filtering, and song mood classification to provide users with relevant and enjoyable song suggestions based on their individual tastes and emotional states.
References
[1]. Y. B. a. A. C. Ian Goodfellow, Deep Learning, 2016.
[2]. "Facial Expression Recognition using Convolutional Neural Networks: State of the Art," 9 Dec 2016.
[3]. A. H. M. P. Timo Ahonen, "Face description with local binary patterns: application to face recognition".
[4]. N. D. a. B. Triggs, "Histograms of Oriented Gradients for Human Detection".
[5]. M. Athavle, "Music Recommendation Based on Face Emotion Recognition," June 2021.
[6]. D. Y. T. Kun Han, "Speech Emotion Recognition Using Deep Neural Network and Extreme Learning Machine," September 2014.
[7]. B. Logan, "Mel Frequency Cepstral Coefficients for Music Modeling," November 2000.
[8]. M. A. A. A. R. Hael Al-bashiri, "Collaborative Filtering Recommender System: Overview and Challenges," Journal of Computational and Theoretical Nanoscience 23(9):9045-9049, September 2017.
[9]. B. M. Raffel‡, "librosa: Audio and Music Signal Analysis in Python," January 2015.
[10]. H. Z. D. Markus Schedl, "Current Challenges and Visions in Music Recommender Systems Research," June 2018.
[11]. X. Jia, "Music Emotion Classification Method Based on Deep Learning and Improved Attention Mechanism," 2022.
[12]. X. Jia, "A Music Emotion Classification Model Based on the Improved Convolutional Neural Network," Hindawi, 2022.
[13]. Dong Liu, "Multi-Modal Fusion Emotion Recognition Method of Speech Expression Based on Deep Learning," Frontiers, 09 July 2021.
[14]. M. NUZZOLO, Music Mood Classification.