The Taro Leaf Blight Detection Project

The TLBProject is an AI research initiative for the early detection of the Phytophthora Colocasiae disease in the Taro (Cocoyam) crop.
The TLBProject is funded by the Responsible Artificial Intelligence Network for Climate Action in Africa (RAINCA).

Research Overview

Key project milestones from the TLBProject.

Android Application for small-holder farmers

We developed an android application for the early detection of the TLB disease. Click on the video to watch a demo of the app.

The first of its kind Taro Leaf Blight Image Dataset

Across Nigeria and Ghana, we collected over ten thousand images of the Taro crop in various stages of the disease, soon to be published.

Open Source Code

The project code is open source and available here.

Read about the research

Each entry summarizes a key aspect of the project.

Data Collection across Nigeria and Ghana.

The TLB Project executed the first ever large scale effort to collect over ten thousand images of the Taro crop in various stages of the Taro Leaf Blight Disease. This entry summarizes our efforts and results.

Training Image Classification and Object Detection Models for Early Detection.

To facillitate the effective early detection of the Taro Leaf Blight disease, the TLB Project training state of the art image classification and object detection models.

Design and Development of an Android Application for Small-Holder Farmers.

The ultimate goal of the TLB Project is to empower small-holder farmers with a solution to enable detect the Taro Leaf Blight disease early. In this entry we describe the features of the android app as well our empowerment efforts for the small-holder farmers.