[M-IS20] Ocean Plastics, an earth science perspective

[MIS20-06] Estimation of plastic litter abundance in cities using big data associated with citizen science and deep learning

*Shinichiro Kako¹, Shintaro Kanno², Atsuhiko Isobe³, Masato Kuki⁴, Kotaro Yamamoto⁴, Fujio Kojima⁴ (1.Graduate School of Science and Engineering, Kagoshima University, 2.Faculty of engineering, Kagoshima University, 3.Research Institute for Applied Mechanics, Kyushu University, 4.Pirika, Inc. / Prika association)

Keywords:citizen science, deep learning, smartphone app “Pirika” , big data, plastic litter abundance in cities

In this study, we constructed a deep learning model using training data consisting of litter images in cities collected by citizen science using a smartphone application “Pirika.” Our deep learning model can classify and quantify the amount of litter in cities into various categories (plastic bags, plastic bottles, cans, and others). The model’s accuracies of object detection and classification were 83.0% and 82.9%, respectively. However, these accuracies of the model are guaranteed only under limited conditions and must be improved for practical use. The results of the learning curve shows that there is a limit to learning with the current training data and raises a concern that the model overfits the training data (validation accuracy of 70% and training accuracy of 99%). In future studies, the versatility of the model must be improved by increasing the size of the learning dataset by including various patterns for detection and classification. The smartphone app “Pirika” must also be updated with a function that disallows taking images that might lower the accuracy of our model.