Trash Image Classification Using Autoencoder

With the expanding sum of waste produced around the world, it has ended up vital to create effective strategies for waste management. One of the key challenges in waste management is the classification of waste based on its sort, because it is time-consuming and requires critical exertion. In this paper, an autoencoder-based deep learning method for classifying trash images was provided. The proposed strategy includes the use of an autoencoder for highlight extraction and classification. Particularly, an autoencoder on a dataset of TrashNet was prepared to memorize the fundamental highlights and after that the learned highlights were used to classify the waste pictures. The proposed strategy was implemented on a dataset of real-world waste pictures and appear that it accomplishes high exactness in classifying diverse sorts of waste. The proposed method has the potential to be conveyed in waste management frameworks to computerize the method of waste classification, in this way sparing time and exertion. The proposed autoencoder-based deep learning method for classifying trash images using TrashNet dataset achieved high accuracy in identifying various types of waste. This approach has the potential to be deployed in trash management systems to automate the process of trash sorting and reduce the time and effort required for it. Overall, this method could prove to be a valuable tool in addressing the challenges of waste management.