Ro'ya is an Arabic word that means vision to indicate the focus of our community on computer vision datasets, methods and applications that impact our African communities.
We aspire to have a diverse community whose members span the whole continent and come from every country in Africa inspired by DLIndaba-X. One of our main beliefs inspired by the DLIndaba community is that “We learn together to build together”. We want to encourage more students and researchers to learn about computer vision fundamentals and its applications which impact Africans. We aim to solve real-world problems that are happening in developing countries, especially the African continent, instead of the current CV research that is focused on developed countries and big tech companies' goals. Our community focus is on a big umbrella of multiple computer vision applications including medical image processing, precision agriculture, remote sensing, traffic monitoring, wildlife monitoring, climate change, AI fairness and others.
To build and facilitate a community of Computer Vision researchers, connect and grow it, spurring and sharing further research, build helpful tools for applications in medicine, agriculture, science and education, to benefit our African communities. We aspire to have a computer vision community with members from every African country and help establish diverse goals that serves the whole continent without bias to certain areas.
To build data sets and tools to facilitate Computer Vision research for Africa, and to decolonize the current Computer Vision research that is mostly focused on Big Tech companies and developed countries goals instead of our needs.
Our community members are helping in multiple initiatives
CV4Africa Workshop in DLIndaba 2023, Ghana. We are working on organizing a computer vision for Africa workshop with two tracks (Datasets - Methods). Where the first track focuses on datasets for Africans on applications that are relevant to us, while the second track is about the algorithms, architectures and computer vision approaches that serve our African communities. We are also working with Zindi to host a challenge as part of the workshop.
Computer Vision for Africa, A Bibliometric Study. We are currently working together on building a bibliometric study for both datasets and methods that are for Africans and by Africans. The goal of this study is to understand the current CV research ongoing in Africa and understand the gaps where we need the community to contribute to.
Weekly Reading/Study Group on Sundays 3:00 pm GMT. We have studied the fundamentals of meta-learning, fewshot learning, and vision transformers. Additionally, we had one hands-on practical on ViTs. Next we are focusing on medical image processing application and learning about generative modelling along with its intersection with few-shot learning.