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{202297,
        author = {UMA SHANTHI S and Dr.K.PADMA PRIYA},
        title = {A Comparative Benchmark of Deep Learning Models and Deployment of a Web Application for Automated Detection of Lumpy Skin Disease in Cattle},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {12},
        pages = {7937-7945},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=202297},
        abstract = {A viral illness called Lumpy Skin Disease (LSD) has a major impact on cow herds all over the world and causes considerable financial losses for the livestock industry. Controlling the disease's spread and lessening its effects require an early and precise diagnosis. Three Convolutional Neural Network (CNN) architectures—VGG19, MobileNetV2, and MobileNetV3—are compared in this study for automatic LSD identification from cattle skin photos. The best model is then implemented as a web-based diagnostic tool. To enhance model generalization, a dataset of 717 cow skin photos from two classes (Lumpy Skin and Normal Skin) was gathered, preprocessed, and enhanced. To improve classification performance, transfer learning was used with pre-trained ImageNet weights. Performance criteria such as accuracy, precision, recall, F1-score, and computing efficiency were used to assess the models. According to experimental data, VGG19 and MobileNetV3 obtained the maximum validation accuracy of 97.18%, with corresponding F1-scores of 0.97 and 0.96. With a model size of 51.21 MB and an inference time of 134 ms—roughly 7.6 times faster than VGG19—MobileNetV3, on the other hand, showed noticeably higher computational efficiency. MobileNetV3 was chosen for deployment because of its balanced accuracy and efficiency performance. A Gradio-based online application that offers users actionable recommendations, confidence scores, and real-time predictions was incorporated with the trained model. The suggested system successfully closes the gap between cutting-edge deep learning research and useful veterinary applications, facilitating early disease management in the cattle sector and enabling quick preliminary screening of LSD in settings with limited resources.},
        keywords = {Lumpy Skin Disease, Deep Learning, Convolutional Neural Networks, MobileNetV3, Transfer Learning, Veterinary Diagnostics, Web Deployment, Gradio},
        month = {May},
        }