Welcome to the Plant Disease Detection Generalization repository! Here, we embark on a journey to revolutionize agriculture through the power of deep learning. Our project is dedicated to creating a robust, accessible, and efficient solution for identifying plant diseases from leaf images. With a focus on leveraging state-of-the-art convolutional neural network (CNN) architectures and transfer learning techniques, we aim to empower farmers worldwide with a tool that enables early disease detection, thus contributing to sustainable agriculture and food security.
Inside this repository, you'll find a treasure trove of resources meticulously crafted to support our mission:
Dive into our collection of Jupyter notebooks, each dedicated to implementing and experimenting with different CNN architectures. Whether it's AlexNet, VGG-16, VGG-19, ResNet, DenseNet, EfficientNet, or ConvNextLarge, we've got you covered. Additionally, explore the notebook for the final model evaluation and comparison to see our journey unfold.
Notebooks-
- AlexNet- https://www.kaggle.com/code/gyanbardhan/alexnet
- VGG16- https://www.kaggle.com/code/gyanbardhan/project-vgg16-2
- VGG19- https://www.kaggle.com/code/clay108/vgg19
- DenseNet- https://www.kaggle.com/code/clay108/project1-densenet121
- ResNet- https://www.kaggle.com/code/gyanbardhan/project1-resnet
- EfficientNet- https://www.kaggle.com/code/clay108/project1-efficientnetb5
- ConvNextLarge- https://www.kaggle.com/code/gyanbardhan/project1-convnextlarge
- Final Model- https://www.kaggle.com/code/clay108/evaluation-of-vgg16-19-an
Explore the datasets that fuel our models' training and validation. Our curated dataset includes preprocessed images meticulously prepared to enhance model performance and accuracy. Test the robustness of our models with the unseen images provided in the test dataset.
- Training Dataset:- https://www.kaggle.com/datasets/gyanbardhan/clahe-plant-disease
- Testing Dataset:- https://www.kaggle.com/datasets/gyanbardhan/groundnuttest
Witness the culmination of our efforts in the form of trained model files ready for deployment and inference. Each model represents hours of fine-tuning and optimization to ensure top-notch performance in plant disease detection.
- AlexNet- https://www.kaggle.com/datasets/gyanbardhan/alexnet123
- VGG16- https://www.kaggle.com/datasets/gyanbardhan/vgg16
- VGG19- https://www.kaggle.com/datasets/clay108/vgg19-123
Experience the magic of our plant disease detection system firsthand with our user-friendly web application. Upload leaf images effortlessly and receive instant disease predictions powered by our trained models.
Delve into the wealth of knowledge that informed our project's methodology and approach. Explore research papers, articles, and resources that shaped our understanding and guided our decisions. ** Link given in Documantation
Access essential links to our deployed web application, notebooks, datasets, and trained models for seamless navigation and exploration.