Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

A 1D-CNN with advanced data augmentation for robust speech emotion recognition.

Neha Chourasia¹, Chhattar Singh Lamba¹, Amit Kumar Gupta²

¹Department of Computer Science and Engineering, Manipal University Jaipur, Jaipur, Rajasthan, India.

Scientific Reports

|June 17, 2026

Summary

This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Bridging Evidence and Policy: Reflections on the Economic Burden of Cancer Care in Ethiopia.

Value in health regional issues·2026

Same author

From pilots to platforms: making digital health work for chronic disease management.

Frontiers in public health·2026

Same author

Development and characterization of triazole-based WDR5 inhibitors for the treatment of glioblastoma.

JCI insight·2026

Same author

Solid Organ Transplantation in Inborn Errors of Metabolism: An Organ-Based, Practice-Oriented Review.

Indian journal of pediatrics·2026

Same author

Consensus on the management of hypertension in individuals with diabetes by Asian-Pacific Society of hypertension (APSH) & Diabetes Asia Study Group (DASG).

Journal of human hypertension·2026

Same author

Hybrid Fusion Models for Parotid Tumors: Strengths, Gaps and Next Steps.

Ultrasound in medicine & biology·2026

Same journal

MT-MRI for detection of renal interstitial fibrosis in renovascular disease.

Scientific reports·2026

Same journal

Detection of underground objects from GPR data using a lightweight YOLO-based approach.

Scientific reports·2026

Same journal

Early systemic inflammatory-metabolic trajectory phenotypes are associated with survival outcomes in metastatic renal cell carcinoma treated with nivolumab.

Scientific reports·2026

Same journal

Water balance components in a dry-seeded rice-wheat system: Untangling the effects of tillage and mulching practices.

Scientific reports·2026

Same journal

Topological approaches to quantum tensor train compression via ZX-calculus and SVD.

Scientific reports·2026

Same journal

determinants of flood impacts and adaptive capacity among market vendors in Walukuba-Masese, Jinja city, Uganda.

Scientific reports·2026

See all related articles

This study introduces a deep learning framework for accurate Speech Emotion Recognition (SER). The model achieves 94.91% accuracy, enhancing human-computer interaction through robust emotion detection from speech.

Area of Science:

Artificial Intelligence
Speech Processing
Machine Learning

Background:

Speech Emotion Recognition (SER) is crucial for intelligent systems in human-computer interaction.
Traditional SER methods struggle with dataset imbalance, speaker variability, and noise, limiting generalization.
Accurate SER is vital for applications in mental health, healthcare, and virtual assistants.

Purpose of the Study:

To propose a robust Conv1D-based deep learning framework for Speech Emotion Recognition (SER).
To enhance SER performance by utilizing complementary acoustic features and data augmentation.
To develop an effective and computationally efficient SER solution for real-world applications.

Main Methods:

Utilized Mel-Frequency Cepstral Coefficients (MFCCs), Zero-Crossing Rate (ZCR), and Root Mean Square Energy (RMSE) for feature extraction.

Keywords:

1D-CNN Data augmentation MFCC RMSE Speech emotion recognition ZCR

Related Experiment Videos

Employed a Conv1D deep learning architecture for hierarchical temporal feature learning from sequential acoustic features.

Applied data augmentation techniques (noise injection, time shifting, pitch modification, time stretching) to improve robustness and class balance.

Main Results:

Achieved a test accuracy of 94.91% and a Macro F1-score of 0.94 across seven emotional categories.
Demonstrated strong recognition capability across diverse emotional speech conditions using a combined multi-dataset.
Validated the effectiveness of integrating discriminative acoustic features, sequential representation, and Conv1D temporal learning.

Conclusions:

The proposed Conv1D-based framework offers a significant improvement in Speech Emotion Recognition performance.
The integration of complementary acoustic features and data augmentation enhances model robustness and generalization.
The framework presents a computationally efficient and effective solution for practical SER applications.