Published Paper Details:

This study evaluates various feature selection techniques for Music Emotion Recognition (MER) by leveraging Mel Frequency Cepstral Coefficients (MFCC) and implementing both Support Vector Machines (SVM) and Convolutional Neural Networks (CNN). Experiments were carried out using a labeled dataset of music samples categorized into five distinct emotions. The impact of various MFCC configurations on SVM-based and CNN-based MER performance was analyzed. Results provide insights into optimal MFCC parameter selection for improved accuracy in MER systems. This research contributes to advancing the field of MER and provides guidelines for enhancing emotion classification in music. Furthermore, the proposed research demonstrates the importance of considering the emotional nuances present in music by utilizing a diverse dataset with multiple emotion categories. By encompassing emotions such as Devotional, Happy, Romantic, Party, and Sad, our study captures a wide range of emotional states expressed through music. This comprehensive approach enables a more thorough understanding of the complexities involved in music emotion recognition and enhances the applicability of the findings in real-world scenarios. The results of this research can lay the groundwork for creating more precise and resilient MER systems, which could enhance fields such as music recommendation, affective computing, and interactive music experiences