Paper Title
Plantd - Detection of Foliar Diseases of Tomato Plant

Abstract
India being an agricultural country, it is the moral responsibility of citizens to take care of the crops. Firstly, it is to be ensured whether the yield is good or not; secondly, whether the crop is prone to diseases; and finally, as well as the preventive measures. It would be very helpful for farmers as well as Phytopathologists if diseases could be detected at an early stage using AI. A thorough study was done on different Convolutional Neural Network (CNN) architecture - ResNet152, InceptionResNetV2, DenseNet201, and InceptionV3 on a dataset of 11,000, (10,000 for training and 1000 for testing) tomato leaf images to classify the diseases out of the 10 classes (healthy and various 9 groups of diseased leaves). DenseNet201 excelled in detecting tomato leaf diseases with an accuracy of 97.6% for ten-class classification. The model is integrated with a very user-friendly UI made using flutter. So, it can be said that deep architecture outperformed shallow designs when it comes to classification of diseases. Keywords - Neural Networks, CNN, AI, Tomato Plant, Disease Detection, Intuitive UI, Flutter, DenseNet201