SDG Impact Award 2025 for Maya Guido
Congratulations! Maya Guido was awarded the SDG Impact Award at the Faculty of Business, Economics and Informatics’ graduation ceremony.
Maya Guido explored in her Bachelor’s thesis the development of an accessible audio annotation tool for visually impaired data annotators.
With her Bachelor's thesis, “Towards inclusivity in machine learning development: Building data annotation tools for visually impaired people”, Maya Guido makes a decisive contribution to the implementation of the Sustainable Development Goals SDG 8 “Decent Work and Economic Growth”, SDG 9 “Industry, Innovation and Infrastructure” and SDG 10 “Reduced Inequalities”.
The Faculty of Business, Economics and Informatics honors outstanding theses that address one or more of the 17 UN Sustainable Development Goals (SDGs).
To build AI that works for everyone, we need everyone to build it (read Abstract)
Many machine learning models feed on data that is traditionally labeled by humans. These human inputs can guide how AI systems understand the world and make decisions. Consequently, AI can inherit biased or narrow perspectives from the crowd that labels its data. This can lead to inequitable outcomes, especially for underrepresented communities who have historically had fewer opportunities to take part in such stages of the ML development pipeline. In her thesis, Maya Guido addresses this challenge by developing a tool that enables blind and visually impaired individuals to actively contribute to data labelling. Built with and for this community, the tool enables users to annotate audio data non-visually leveraging screen reader output and accessible keyboard navigation. Inclusive machine learning tools like these are a step towards ensuring that AI is shaped by the diversity of those it is meant to serve.