Recycle Items Classification for Waste Management using CNN Models

The rising amount of waste produced worldwide presents serious environmental concerns, and there is a need to create automated waste classification systems to enable effective recycling processes. This study proposes a deep learning-based method employing Convolutional Neural Networks (CNNs) for waste classification into pre-defined categories like cardboard, compost, glass, metal, paper, plastic, and trash. The system we suggest takes gain of many pre trained CNN models MobileNetV2 Resnet 50 EfficientNetB0 they are trained on finetuned different wastes dataset ensemble model created to make this classification accuracy pool by individual strength of each model. Many experiments find proof that the model ability to get high accuracy in wastes categorization and pinpoints its practical usage in automate waste systems research adds a development to increase sustained and environment prevention. The proposed ensemble architecture yields the better when compared with the base models and other existing models.