TY  - JOUR
AU  - Tiwari, Shreya
AU  - Kumar, Devansh
AU  - Mahajan, Akshit
AU  - Sachar, Silky
PY  - 2025
DA  - 2025/09/14
TI  - Emotion Detection from Speech Using CNN-BiLSTM with Feature Rich Audio Inputs
JO  - ICCK Transactions on Machine Intelligence
T2  - ICCK Transactions on Machine Intelligence
JF  - ICCK Transactions on Machine Intelligence
VL  - 1
IS  - 2
SP  - 80
EP  - 89
DO  - 10.62762/TMI.2025.306750
UR  - https://www.icck.org/article/abs/TMI.2025.306750
KW  - speech emotion recognition
KW  - deep learning
KW  - CNN-BiLSTM
KW  - RAVDESS
KW  - MFCC
KW  - real-time prediction
KW  - human-computer interaction
KW  - audio processing
KW  - web deployment
KW  - affective computing
AB  - In the age of increasing machine-mediated communication, the ability to detect emotional nuances in speech has become a critical competency for intelligent systems. This paper presents a robust Speech Emotion Recognition (SER) framework that integrates a hybrid deep learning architecture with a real-time web-based inference interface. Utilizing the RAVDESS dataset, the proposed pipeline encompasses comprehensive preprocessing, data augmentation techniques, and feature extraction based on Mel-Frequency Cepstral Coefficients (MFCCs), Chroma features, and Mel-spectrograms. A comparative experiment was run against a standard machine learning classifier such as K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Random Forest, and XGBoost. The experimental results indicate that the CNN-BiLSTM-Conv1D model proposed is much better as compared to conventional models with a state-of-the-art classification accuracy of 94%. The model was further evaluated using ROC-AUC curves and per-class performance metrics. It was subsequently deployed using a Flask-based web interface that enables users to upload voice inputs and receive real-time emotion predictions. This end-to-end system addresses the shortcomings of earlier SER approaches---such as limited temporal modeling and reduced generalization---and showcases practical applicability in domains like mental health monitoring, virtual assistants, and affective computing.
SN  - 3068-7403
PB  - Institute of Central Computation and Knowledge
LA  - English
ER  - 

@article{Tiwari2025Emotion,
  author = {Shreya Tiwari and Devansh Kumar and Akshit Mahajan and Silky Sachar},
  title = {Emotion Detection from Speech Using CNN-BiLSTM with Feature Rich Audio Inputs},
  journal = {ICCK Transactions on Machine Intelligence},
  year = {2025},
  volume = {1},
  number = {2},
  pages = {80-89},
  doi = {10.62762/TMI.2025.306750},
  url = {https://www.icck.org/article/abs/TMI.2025.306750},
  abstract = {In the age of increasing machine-mediated communication, the ability to detect emotional nuances in speech has become a critical competency for intelligent systems. This paper presents a robust Speech Emotion Recognition (SER) framework that integrates a hybrid deep learning architecture with a real-time web-based inference interface. Utilizing the RAVDESS dataset, the proposed pipeline encompasses comprehensive preprocessing, data augmentation techniques, and feature extraction based on Mel-Frequency Cepstral Coefficients (MFCCs), Chroma features, and Mel-spectrograms. A comparative experiment was run against a standard machine learning classifier such as K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Random Forest, and XGBoost. The experimental results indicate that the CNN-BiLSTM-Conv1D model proposed is much better as compared to conventional models with a state-of-the-art classification accuracy of 94\%. The model was further evaluated using ROC-AUC curves and per-class performance metrics. It was subsequently deployed using a Flask-based web interface that enables users to upload voice inputs and receive real-time emotion predictions. This end-to-end system addresses the shortcomings of earlier SER approaches---such as limited temporal modeling and reduced generalization---and showcases practical applicability in domains like mental health monitoring, virtual assistants, and affective computing.},
  keywords = {speech emotion recognition, deep learning, CNN-BiLSTM, RAVDESS, MFCC, real-time prediction, human-computer interaction, audio processing, web deployment, affective computing},
  issn = {3068-7403},
  publisher = {Institute of Central Computation and Knowledge}
}