Waste management techniques should be improved to reduce environmental pollution. Problematic situations arise while plastic pollution occurs. The challenging task is to collect plastic waste from the environment. In this era, computer vision helps solving real-life problems. Environmental problems can be sorted out using the advanced technology associated with computer vision. Using deep learning models, different types of food waste and plastic can be identified. In this research work, an automated waste separator bin is implemented using deep learning models and cloud computing. After going through the experimental process with different models, the YOLOv8 object detection model achieves the best mean average precision (mAP50) of 99%. The automated waste separator bin has a multi- box system for collecting categorical waste. Using this intelligent system, some recyclable waste, like plastic, can be sorted out and reduced from mixing with the food waste and other categorical waste. Waste management will be improved by the ability to separate plastic and food waste with this automatic bin at the initial stage of waste collection.