查看文章

Traditionally, IoT has been used independently to create smart systems for the betterment of human society. The use of too many sensors either complicates it or increases the dependence of sensors on each other which becomes a problem in case a sensor becomes faulty. Smart dustbins have also been developed traditionally using only IoT and cloud technology which only sends alerts when the dustbin is full. However, the main task of waste segregation is still done manually which needs to be automated. Our proposed work consists of a unique IoT-based waste segregation system that uses deep learning for waste segregation, reducing the use of sensors and making it easy to install and set up. Our proposed work performs accurate waste classification and also alerts when the dustbin is full using cloud technology for faster alerts. In this paper, transfer learning techniques were used to train the pre …