Deep Neural Garbage Recognition: An Augmented Reality Study Case

Diogo Fernandes Costa Silva, Arthur Ricardo de Sousa Vitória, Arlindo Rodrigues Galvão Filho

Abstract: The recyclability of waste is essential in modern society. Rapid industrialization and urbanization have led to increased waste generation, presenting significant challenges for efficient urban waste management. In this context, this work combines augmented reality with deep learning models to detect, recognize and segment waste in a scene. As a case study, this work conducts experiments using images of waste collected from home, streets, and public spaces. The experiment results show that deep learning models can assist waste recycling in augmented reality scenarios, contributing to sustainable industry practices by enabling more efficient sorting, reducing contamination in recyclable streams, and promoting resource conservation through innovative technological integration.

External IDs:dblp:conf/aivr/SilvaVF25