Multi-Class Waste Classification Image Dataset for CNN Training and TFLite Deployment on Edge Devices

This dataset contains more that 100,050 labeled images across five waste categories: Recyclable, Non-Recyclable, E-Waste, Organic, and StoreDropOff materials. It was created to support the training and evaluation of convolutional neural networks (CNNs) for smart waste classification systems. The dataset is specifically tailored for use with lightweight architectures (e.g., EfficientNetB0, MobileNetV2) and deployment on edge AI platforms using TensorFlow Lite. This data was used in a study exploring the performance of deep learning models under resource-constrained hardware for real-time classification tasks in waste management. Ideal for sustainability-focused AI research and embedded ML development.