A Hybrid ResNet50 and GoogLeNet Deep Learning Architecture for Early Detection of Tomato Leaf Diseases

Author(s)	Mr. Ganesh Sarkar, Prof. Praveen Kumar Gautam, Dr. Dilip Singh Solanki
Country	India
Abstract	. This paper extends that study by constructing a hybrid architecture that fuses both backbones at the feature level: the final global-average-pooling activations of ResNet50 (2048-d) and GoogLeNet (1024-d) are concatenated into a single 3072-d representation and passed through a new classification head. Training reached a best recorded validation accuracy of 99.60%, and confusion-matrix evaluation of the exported network over the full 1,500-image validation set correctly classified 1,491 images (99.60% accuracy; macro-averaged precision, recall, and F1-score of 99.60%), reducing total misclassifications from 13 (ResNet50) and 18 (GoogLeNet) to 9.
Keywords	Tomato Leaf Disease, Deep Learning, Hybrid CNN, Feature Fusion, ResNet50, GoogLeNet, Transfer Learning, Image Augmentation, Plant Pathology, Precision Agriculture, MATLAB
Field	Engineering
Published In	Volume 8, Issue 5, September-October 2026
Published On	2026-10-12

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9687-828-838	editor@ijfmr.com

International Journal For Multidisciplinary Research