Waste Image Dataset – Ndiawdoune, Senegal and Reorganized Kaggle Waste Data

This dataset was created as part of a master's thesis focused on evaluating the performance of state-of-the-art deep learning models in computer vision on local African data. It consists of two subsets: 1. A dataset collected in the village of Ndiawdoune, Saint-Louis, Senegal, containing labeled images of waste in natural environments.2. A reorganized version of the publicly available Kaggle Waste Classification dataset, restructured to match the format and task of the local dataset. The goal is to enable fair comparative evaluation of EfficientNet-B7, ViT-B/16, and ViM-B models on both local and public data. This resource aims to support research on the adaptability of computer vision models to African contexts. 🔗 GitHub repository containing code and documentation