Child Growth Monitor

NGO | Computer Vision | Deep Learning | Augmented Reality

Problem:

Hunger and malnutrition are not a lack of food only, it is a more complex health issue. Child malnutrition is a global problem. Parents and frontline workers often don't know that the child is malnourished and take measures too late. Accurate data on the nutritional status of children is unreliable or non-existent. This hampers a determined response by emergency workers and policymakers.

The required measurements are done by a wooden board (body height), by scales (body weight), and by tape (mid-upper arm circumference). Measuring children with these devices is labor-intensive and often requiring two or more staff members. Also, they require physical contact and are known to be prone to errors leading to invalid measurements.

Solution

CGM is a project of Welthungerhilfe, a German NGO working in the fields of development cooperation and emergency aid. We are introducing a new technology to detect malnutrition: A body scan is taken with a smartphone that has a depth sensor. The solution is based on a mobile app using augmented reality in combination with artificial intelligence. By determining weight and height through a scan of children, the app provides a quick and touchless way to measure children and detect early warning signs of malnutrition.

For diagnosing malnutrition, Welthungerhilfe is interested mainly in the weight, height, and middle-upper arm circumference. According to the World Health Organization, the combination of these three measurements with age and sex of a child enables a diagnosis of malnutrition.

We created a product concept to ensure the AI's predictions are trustworthy and perform consistently well. We wrote this Medium article about this reliability concept.

My Role

The Welthungerhilfe is a German NGO which runs the Child Growth Monitor as an innovation project. I work in the project as Senior ML Engineer. I develop new ML models to accurately and reliably estimate the body height and weight of a child. Find the code on GitHub and have a look at the project website.

Technologies:

Tensorflow, Python, AzureML

Back to Projects