Biodegradable and Non-biodegradable Materials Segregation Using AI-enabled Mobile Application

Biodegradable materials are those materials that can be easily broken down into simpler decomposable units by the various microorganisms, whereas non-biodegradable substances include those materials that do not decompose easily, contributing to environmental pollution. In this paper, a novel method has been proposed for the categorisation of biodegradable and non-biodegradable materials using deep learning model. In this work, a MobileNetV2 model augmented with multi-head attention (MHA) mechanism has been used for the categorisation of biodegradable and non-biodegradable products. The proposed model is deployed in a mobile application built using Flutter and Dart frameworks. Therefore, the proposed model is termed ‘MHA-Mobile’. The proposed method classifies the biodegradable and non-biodegradable materials with 95.19% accuracy. Thus, the proposed model when deployed in form of mobile app has significant usefulness in biodegradable and non-biodegradable materials segregation.