Affiliation:
1. Graduate School of Science and Technology, Nara Institute of Science and Technology, Ikoma 630-0192, Japan
2. Graduate School and Faculty of Information Science and Electrical Engineering, Kyushu University, Nishi-ku, Fukuoka 819-0395, Japan
Abstract
Litter has become a social problem. To prevent litter, we consider urban planning, the efficient placement of garbage bins, and interventions with litterers. In order to carry out these actions, we need to comprehensively grasp the types and locations of litter in advance. However, with the existing methods, collecting the types and locations of litter is very costly and has low privacy. In this research, we have proposed the conceptual design to estimate the types and locations of litter using only the sensor data from a smartwatch worn by the user. This system can record the types and locations of litter only when a user raps on the litter and picks it up. Also, we have constructed a sound recognition model to estimate the types of litter by using sound sensor data, and we have carried out experiments. We have confirmed that the model built with other people’s data enabled to estimate the F-measure of 80.2% in a noisy environment through the experiment with 12 participants.
Subject
Management, Monitoring, Policy and Law,Renewable Energy, Sustainability and the Environment,Geography, Planning and Development,Building and Construction