Samadisco/PHG-RAI

The Official Repository of the PHG Foundation Africa - Responsible Artificial Intelligence Initiative # PHG-RAI: Responsible AI for African Image Data Welcome to PHG-RAI, a repository dedicated to curating image datasets for responsible AI development in the African context. Our mission is to accelerate the development of models and analysis of image data by data scientists in the sub-region. # Overview PHG-RAI aims to address the lack of diverse and representative image datasets in Africa, hindering the development of AI models that cater to the region's unique needs. We focus on curating high-quality images across four initial categories: 1. Health Images: Medical images, including diseases prevalent in Africa, to support healthcare AI applications. 2. Skin Images: Diverse skin tones and conditions, enabling AI systems to better understand and serve African populations. 3. Plant Images: Various plant species, including crops and trees, to support agricultural AI and environmental monitoring. 4. Animal Images: Wildlife and domestic animals, crucial for conservation, livestock management, and ecological research. # Goals 1. Provide a comprehensive and diverse image dataset for AI model development and testing. 2. Foster collaboration among data scientists, researchers, and organizations working on AI applications in Africa. 3. Promote responsible AI practices, ensuring inclusivity, privacy, and ethical considerations in AI development. # Contributing We welcome contributions from individuals and organizations passionate about responsible AI and African image data. You can contribute by: 1. Submitting image datasets: Share your image collections, following our guidelines and ethical standards. 2. Labeling and annotating images: Help us enrich the datasets with accurate labels and annotations. 3. *Participating in discussions*: Engage with our community to shape the future of PHG-RAI and responsible AI in Africa. *License and Ethics* PHG-RAI operates under the Creative Commons Attribution 4.0 International License (CC BY 4.0). We adhere to ethical guideline …