Copyright © 2026 Authors retain the copyright of this article. This article 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.

@article{196933,
        author = {Nuthalapati Gopi Krishna and Shaik Yasir and Vadlamudi Ramu and Shaik Jaheer Basha and Pachhala Naga Babu},
        title = {Malware Detection and Classification Using ResNet-50},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {5473-5477},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=196933},
        abstract = {Aims: To develop and evaluate a deep learning-based malware detection and classification system using the ResNet-50 architecture that converts malware binary files into image representations for automated analysis, overcoming the limitations of traditional signature-based detection methods. Study Design: Experimental deep learning study involving dataset preprocessing, model design and training, performance evaluation, and comparative analysis against traditional detection methods. Place and Duration of Study: Department of Information Technology, Vasireddy Venkatadri Institute of Technology, Guntur, India, 2025–2026. Methodology: Malware binary samples from the Microsoft Malware Classification Challenge dataset were converted into grayscale images and resized to 224×224 pixels. The ResNet-50 architecture—pre-trained on ImageNet and fine-tuned on malware images—was employed for deep feature extraction and classification. The training pipeline included data augmentation, batch normalization, and an Adam optimizer with cross-entropy loss. Performance was assessed using accuracy, precision, recall, and F1-score across nine malware families. Results: The proposed ResNet-50-based system achieved an overall classification accuracy of 92%, with precision of 91.4%, recall of 90.8%, and an F1-score of 91.1% on the held-out test set. The model outperformed traditional signature-based and heuristic approaches, demonstrating superior performance on both known and obfuscated malware variants. Conclusion: The ResNet-50 framework provides an effective, automated, and scalable solution for malware detection and classification. The image-based representation of binary files enables the model to capture complex structural patterns, making it suitable for real-world cybersecurity deployment. Future work should explore hybrid architectures, larger datasets, and real-time integration.},
        keywords = {Malware Detection, ResNet-50, Deep Learning, Binary Visualization, Convolutional Neural Network, Cybersecurity, Image Classification.},
        month = {April},
        }