Marine Litter Detection
MI4People aims to research on a system that can identify ocean plastic and marine litter from satellite images in (almost) real-time. It will be accessible for environmental NPOs, companies, activists, and authorities for no charge. This system will allow better strategic decision making and coordination of ocean cleaning efforts. At the same time, the output of the system (e.g., in form of an interactive map) can be used to strengthen the public attention to the problem of marine litter.
MI4People plans to create a Machine Learning (ML) model that can replicate past and predict future climate change at higher accuracy that the current models. This “black-box” ML model will likely have lower level of explainability in comparison to classical climate models, however, it might potentially deliver substantially higher prediction accuracy. Thus, this approach is expected to be an interesting and valuable complement to existing climate research methods.
Supporting Elderly Inclusion
MI4People plans to research on a system that will automatically identify needs of elderly population based on social media posts, forum threads and web-articles created by this age group. This system will use web-scrapping and Natural Language Processing/Understanding (NLP/NLU) as the main technologies. This project will open a new perspective on needs of elderly people (compared to traditional surveys and subject matter experts' estimations). The results of this research can be used by researchers, companies and authorities to provide better services to and to foster inclusion of aged population. Similar projects might be performed also for other population groups.