The objective of this challenge is to develop robust machine learning models that can accurately predict all diseases present in images of corn, pepper, and tomato crops.

Zindi, a data science platform, launched the Ghana Crop Disease Detection Challenge, which involves crop disease detection using entry-level smartphones.

The challenge is the result of a collaboration with Zindi, GIZ, Digital Transformation Centre Ghana, Fairforward, and the Responsible AI Lab.

According to Zindi, the objective is to create effective machine learning models that can anticipate the presence of diseases in different vegetables in order to combat food insecurity in Sub-Saharan Africa.

Participants are requested to incorporate approaches such as Gradient-Weighted Class Activation Mapping, Local Interpretable Model-Agnostic Explanations, and SHapley Additive exPlanations models.

The techniques, according to the organizers, should provide clear, visual explanations of how the models create their predictions, providing transparency and building trust in AI solutions.

The competition is aimed at African citizens and will be accessible for entries from October 4 until December 15.

The first place winner will get $4000 in cash, followed by $2500 for second place and $1500 for third place.

Zindi emphasised the significance of using technology to combat food insecurity.

It said: “By harnessing the power of machine learning, we aim to develop advanced solutions for detecting and identifying multiple diseases in three vital crops: corn, pepper, and tomatoes.

“The models and solutions developed in this challenge will support accurate and timely disease detection, enhance agricultural productivity and sustainability, and ensure food security for millions of people.”