Combatting Malaria

Solution:

The Humbug project works to monitor the spread of malaria by detecting and identifying different species of mosquito with smartphone technology. The project team required algorithms that could detect and identify different species of mosquitoes using the acoustic signature of their flight tones. This capability needed to be robust across the huge variety of recordings, locations, time of year and background noises, which added an extra layer to the challenge.

Mind Foundry contributed to the HumBug project by providing automated, state-of-the-art machine learning models for the classification of mosquito sounds accelerating the development of new algorithms. The Mind Foundry Platform provides an invaluable tool to experiment with different detection techniques, allowing rapid performance improvement and providing insight into how mosquito wing shape relates to their acoustic signature.

Results:

Working with experts in the Humbug project, Mind Foundry hosted the world’s first mosquito detection and classification competition, further fuelling the algorithm development. The project saw results, including:

• Mosquito sounds recorded by a smartphone could be analysed by an AI detection algorithm.

• 90% accuracy of mosquito classification.

• Performance is faster and more accurate at detecting the presence of certain species and sex of mosquitoes than existing methods.

Click here for more information on the Humbug project.