Music Recommendation Based on Emotion Recognition

Abstract

In an decreasingly digital and data- driven world, personalization of services and happy recommendation is an important aspect to ameliorate the stoner experience. In recent times, there has been a major elaboration in music recommendation system, fastening on stoner sentiment and song characteristics. The proposed system takes advantage of the idea that visual stimulants, similar as the content of images, can be used to infer the stoner\'s emotional state. By recycling facial expressions and other visual cues, the system detects and classifies feelings similar as happiness, sadness, wrathfulness and others. This emotional data is also used to elect music tracks that align with the detected emotional state. The recommendation system frame is displayed by druggies rulings that enable song recommendation on the bases of the stoner\'s preferences or intensity of current feelings. Playing music based on the mood of the user is an application of deep learning which is introduced to the listeners. This can be done by figuring out the user\'s facial expression which depicts their mood.

Introduction

I. INTRODUCTION

The domain of music recommender has undergone substantial development, encompassing an array of algorithms and methodologies crafted to facilitate users in discovering music that aligns with their preferences. Typically, these systems leverage implicit preferences and listening history along with user emotions to produce.

Music plays a vital role in our daily lives and has a deep effect on human emotions. Given the influence of music on our feelings and experiences, there is an increasing need for music recommendation systems that are sensitive to the emotional demands of their users. This study attempts to fill this gap by presenting a deep network-based image processing and emotion identification system for music recommendations. It presents a novel method that considers users' emotional states in real time. The system can suggest music that not only fits the users' musical tastes but also expresses the feelings they are experiencing at the moment.

Because it prioritises emotional understanding over other factors, this signifies a paradigm shift in the way we think about recommending music. Facial expressions are captured and interpreted by the system to identify emotions like surprise, happiness, sadness, and anger.

II. RELATED WORK

One of the most important approaches to music recommendation systems was cooperative filtering, which involves recommending music grounded on a stoner's once harkening history and the listening history of analogous druggies. still, cooperative filtering has limitations in directly prognosticating a stoner's mood and furnishing individualized recommendations grounded on their current emotional state. Facial expression analysis is used in colorful fields similar as psychology, neuroscience, and computer vision to determine a person's emotional state grounded on their facial expressions. Recently, several studies have investigated the use of facial expressions analysis in music recommendation systems. In a study by Yang etal. (2014), music player recommender was developed which uses facial expression analysis to determine a stoner's emotional state and give individualized music recommendations. The system used a deep recognition network to learn the mapping between facial expressions and emotional countries and a cooperative filtering algorithm to induce substantiated music recommendations. In another study by Yang etal.(2016),music recommender was developed that used a multimodal approach incorporating both facial emotion analysis and physiological cues. The system used a deep literacy- grounded model to descry and classify facial expressions and a physiological signal- grounded model to measure the stoner's physiological responses, similar as heart rate and skin conductance. The system also used a mongrel approach of cooperative filtering and content- grounded filtering to induce substantiated music recommendations. In the recent study by Li etal. (2020), developed the music recommendation system that used a facial expression- grounded emotion recognition model to induce substantiated music recommendations.

This system uses a convolutional neural network (CNN) algorithm to describe and classify facial expressions and k- nearest neighbor algorithm to induce substantiated music recommendations grounded on the stoner's emotional states.

III. OBJECTIVES

The proposed system allow the user to represent the interaction between the user and the music player. Providing users with a list of songs directly into their personal playlist based on mood.

1. Real-time capture: In this module, the system is to capture the user's face correctly.
2. Face Recognition: Here it will take user's face as input
3. Emotion Detection: In this section, the features of the user image are extracted to detect the emotions and the system will generate captions according to the user’s emotion.
4. Music Recommendation: A song is suggested by the recommendation module by mapping the user's emotions with the mood type of the song.

IV. METHODOLOGY

Data Collection we collected a dataset of captured emotions of user and music preference conditions from a group of actors. Actors were asked to hear to music when their face emotions were recorded using a camera. They were also asked to rate the music they heard grounded on their emotional response.

1. Facial Expression Analysis We used a deep literacy- grounded model for facial expression analysis. The model had multiple convolutional layers, which were followed by fully linked layers that discovered the relationships between expressions and emotions on the face. The model comported of several convolutional layers followed by completely connected layers that learned mappings between facial emotions and expressions. Emotional countries were classified into several orders, similar as happy, sad, angry, and neutral.
2. Music recommendation We used a cooperative filtering algorithm for music recommendation. The algorithm generates substantiated music recommendations grounded on the stoner's emotional state prognosticated by a facial expression analysis model.
3. Evaluation estimated the performance of the proposed system by measuring its delicacy in prognosticating the stoner's emotional state and its effectiveness in furnishing substantiated music recommendations. We also conducted a stoner study to estimate the effectiveness of the system in furnishing substantiated music recommendations.
4. Results The results indicate that the proposed system can directly prognosticate the stoner's emotional state and give individualized music recommendations that match their mood. The system achieved high delicacy in prognosticating the stoner's emotional state, with an averageF1 score of 0.85.
5. Overall, the proposed system demonstrated the effectiveness of using facial expression analysis and cooperative filtering in a music recommendation system to give individualized recommendations that match the stoner's emotional state. The system has the implicit to ameliorate stoner satisfaction and engagement with music recommendation systems by furnishing more accurate and individualized recommendations.

V. SYSTEM REQUIREMENTS

A. Hardware Requirements

1. System                                               :              Intel i3 with speed of 2.4 GHz.
2. Hard Disk                                          :              120 GB.
3. Monitor                                             :              15’’ LED
4. Input Devices                                    :              Keyboard, Mouse
5. Ram                                                    :              4 GB

???????B. Software Requirements

1. Operating system                                :              Windows 7 64-bit.
2. Coding Language                             :              PYTHON 3.6
3. Tools                                                  :              Anaconda

VI. ACKNOWLEDGEMENT

I would like to express our deep gratitude to all the faculty members of CSE Department for giving valuable and constructive Suggestion during the Planning and development of this project. Their willingness to give their time So generously has been very much appreciated. We would like to thank all the professors of KSIT for their continuous support and encouragement.

References

[1] Ramya Ramanathan, Radha Kumaran, Ram Rohan R, Rajat Gupta, and Vishalakshi Prabhu, an intelligent music player based on emotion recognition, 2nd IEEE International Conference on Computational Systems and Information Technology for Sustainable Solutions 2017. [2] Shlok Gilda, Husain Zafar, Chintan Soni, Kshitija Waghurdekar, Smart music player integrating facial emotion recognition and music mood recommendation, Department of Computer Engineering, Pune Institute of Computer Technology, Pune, India, (IEEE),2017. [3] Deger Ayata, Yusuf Yaslan, and Mustafa E. Kamasak, Emotion-based music recommendation system using wearable physiological sensors, IEEE transactions on consumer electronics, vol. 14, no. 8, May 2018. [4] Ahlam Alrihail, Alaa Alsaedi, Kholood Albalawi, Liyakathunisa Syed, Music recommender system for users based on emotion detection through facial features, Department of Computer Science Taibah University, (DeSE), 2019. [5] Research Prediction Competition, Challenges in representation learning: facial expression recognition challenges, Learn facial expression from an image, (KAGGLE).

Copyright

Copyright © 2023 Poonam ., Srishti Srivastava, Sushma Srikanta Kurandwad, Parvashree H M. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.