By leveraging artificial intelligence and computer vision, we aim to help reduce crop losses and promote sustainable farming practices.

Click below to checkout the website

A school project of utilizing YOLOv5 Object Detection algorithm to train a pre-trained model with and test it against a dataset containing more than 6000 tomato leaves of 5 classes: Healthy, Bacterial Spot, Early Blight, Late Blight, Powdery Mildew.

Hyperparameter-tuned Xception model for plant disease detection with 97.23% accuracy – published at IEEE ICAIQSA 2024

Repositori ini digunakan untuk membuat aplikasi web menggunakan streamlit untuk mendeteksi penyakit pada daun tanaman tomat.

A deep learning-based classifier for detecting diseases in tomato plant leaves using a Convolutional Neural Network (CNN). The project includes data preprocessing, model training, evaluation, and a Streamlit web interface for real-time disease prediction and visualization.

Few-shot and Test-Time Adaptation based Tomato Leaf Disease Detection using Vision Transformers

Built a multi-task deep learning system for tomato leaf disease classification and severity prediction using ResNet50, CBAM, and cross-task attention; improved severity F1 from 0.32 to 0.75 under partial-label conditions.

🍅 Detect tomato leaf diseases using minimal data with a robust framework combining Few-Shot Fine-Tuning and Test-Time Adaptation for real-world applications.