School of Automation and Electrical Engineering, University of Science and Technology Beijing, Beijing 100083, China.
School of Automation, Guangdong University of Petrochemical Technology, Maoming 525000, Guangdong, China; Beijing Advanced Innovation Center for Soft Matter Science and Engineering, Beijing University of Chemical Technology, Beijing 100029, China; Department of Chemistry, Institute of Inorganic and Analytical Chemistry, Goethe-University, 60438 Frankfurt, Germany.
Waste Manag. 2021 May 1;126:247-257. doi: 10.1016/j.wasman.2021.03.017. Epub 2021 Mar 26.
Accurate waste classification is key to successful waste management. However, most current studies have focused exclusively on single-label waste classification from images, which goes against common sense. In this paper, we move beyond single-label waste classification and propose a benchmark for evaluating the multi-label waste classification and localization tasks to advance waste management via deep learning-based methods. We propose a multi-task learning architecture (MTLA) based on a convolutional neural network, which can be used to simultaneously identify and locate wastes in images. The MTLA comprises a backbone network with proposed attention modules, a novel multi-level feature pyramid network, and a group of joint learning multi-task subnets. To achieve joint optimization of waste identification and location, we designed the loss functions according to the concepts of focusing and joint. The proposed MTLA achieved performance similar to that of experts and had high scores for multiple tasks related to waste management. Its F1 score exceeded 95.50% (95.12% to 95.88%, with a 95% confidence interval) on the multi-label waste classification task, and the average precision score was over 81.50% (@IoU = 0.5) on the waste localization task. To improve interpretation, heatmaps were used to visualize the salient features extracted by the MTLA. The proposed MTLA is a promising auxiliary tool that can improve the automation of waste management systems.
准确的垃圾分类是成功进行废物管理的关键。然而，大多数现有的研究都只专注于从图像中单标签的废物分类，这有悖于常理。在本文中，我们超越了单标签废物分类，提出了一个基准，用于评估多标签废物分类和定位任务，以通过基于深度学习的方法推进废物管理。我们提出了一种基于卷积神经网络的多任务学习架构（MTLA），可用于同时识别和定位图像中的废物。MTLA 包括一个带有提出的注意力模块的骨干网络、一个新颖的多级特征金字塔网络和一组联合学习的多任务子网。为了实现废物识别和定位的联合优化，我们根据聚焦和联合的概念设计了损失函数。所提出的 MTLA 的性能与专家相当，并且在与废物管理相关的多个任务中得分很高。在多标签废物分类任务中，其 F1 得分超过 95.50%（置信区间为 95.12%至 95.88%），在废物定位任务中的平均精度得分超过 81.50%（@IoU=0.5）。为了提高解释性，使用热图可视化了 MTLA 提取的显著特征。所提出的 MTLA 是一种很有前途的辅助工具，可以提高废物管理系统的自动化程度。
