Nigerian Indigenous Food Ingredient Image Dataset (NIFID)

The Nigerian Indigenous Food Ingredient Image Dataset (NIFID) contains 4,306 photographs of 16 indigenous Nigerian food ingredient classes collected from real-world environments including open markets, home kitchens, and rural settings across Nigeria.

The dataset was originally collected in connection with the Deep Nutrivision project (Uzo et al., 2025, ICDIA) and is released here as a standalone, cleaned, and verified archive to support reproducible research in food recognition, nutrition informatics, cultural informatics, and machine learning for underrepresented food environments.

Classes (16): Abacha, Black Beans, Fio Fio, Groundnut, Ogbono, Okpa, Okro, Onions, Peeled Egusi, Processed Ukwa, Sweet Potato, Tomatoes, Ugu Leaf, Waterleaf, White Yam, Yellow Corn.

Image properties: JPEG/PNG format, variable resolution (mobile-captured), uncontrolled lighting and backgrounds.

Cleaning note: The initial collection contained 4,310 files. Four corrupt image files were identified and removed during quality control. This archived release contains 4,306 verified usable images.