Efficient Waste Sorting for Circular Economy - A Confidence-guided comparison between One-Vs-All and One-Vs-Rest Classification Strategies with Human-in-the-Loop for Automated Waste Sorting

Sep 21, 2026

Uploaded by Ivy

The work addresses how complex, country-specific waste-disposal regulations hinder residents and slow the transition to a Circular Economy, especially in Germany where household waste sorting varies by municipality. It evaluates two deep-learning multi-class classification strategies, One-vs-All (OvA) and One-vs-Rest (OvR), using a dataset aligned with Goslar’s waste categories. Beyond overall accuracy, it studies misclassification-prone samples via confidence thresholds for human review, balancing annotation effort against errors to enable configurable AI sorting tools.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages8

Reading Time20 min

AccessFree to download

Abstract

Index Terms

I. Introduction

Efficient Waste Sorting for Circular Economy: A Confidence-guided comparison between One-Vs-All and One-Vs-Rest Classification Strategies with Human-in-the-Loop for Automated Waste Sorting Mohammed Fahad Ali, Dominique Briechle, Marit Briechle-Mathiszig, Tobias Geger and Andreas Rausch Institute for Software and Systems Engineering Clausthal University of Technology Germany e-mail: {mohammed.fahad.ali, dominique.fabio.briechle, marit.elke.anke.mathiszig, thomas.tobias.marcello.geger, [andreas.rausch](andreas.rausch}@tu-clausthal.de)[}](andreas.rausch}@tu-clausthal.de)[@tu-clausthal.de](andreas.rausch}@tu-clausthal.de) arXiv :2607 .02230v 1 [ cs .CV] 2 Jul 2026 Abstract—The complexity of waste disposal regulations across European countries poses significant challenges for the residents and hinders the transition to a Circular Economy. In Germany, the proper sorting and disposal of household waste remains challenging across municipalities. Consequently, substantially reducing incorrectly disposed waste is vital for improving waste management and advancing the Circular Economy. Artificial Intelligence (AI)-based waste sorting solutions can support residents through user-friendly tools, such as mobile applications, that guide proper waste disposal. To be effective in supporting the Circular Economy, however, these solutions must be configurable to reflect the specific waste sorting scheme of individual municipalities in Germany. In the scope of this work, an evaluation and analysis are performed of two prominent classification strategies: One-vs-All (OvA) and One-vs-Rest (OvR). The research uses adataset constructed in alignment with the waste categories and sorting scheme of the city of Goslar in Germany. Moreover, this work aims to extend beyond the overall performance by examining the behavior of OvA and OvR classification strategies in identifying samples likely to be misclassified. These classification strategies are compared by applying varying confidence thresholds to identify uncertain samples for subsequent human review. This evaluation aims to balance the number of misclassifications against the human effort required for data annotation. Index Terms—Circular Economy, Multi-class Classification, Deep Learning, One-Vs-Rest Classification, Waste Sorting. I. INTRODUCTION Climate change is one of the major human-driven challenges of the 21st century [23] . In recognition of its importance, the United Nations 2030 agenda includes sustainable consumption and production as one of the 17 Sustainable Development Goals [22] . In this context, the Circular Economy framework aims for a more sustainable approach by keeping products inuse for as long as possible. Unlike its linear counterpart, it focuses on reusing resources in continuous cycles to support sustainable outcomes [24] . In a Circular Economy, effective waste management is vital for maintaining material circulation and reducing resource extraction, and depends partly on proper household waste sorting. Consequently, correctly sorting waste is a structural prerequisite for the Circular Economy, as This work has been developed in the project “6RLogistics”(Research Grant Number: 03EI5025B) and is funded by the Federal Ministry for Economic Affairs and Energy (BMWE), Germany. incorrect sorting contaminates recyclables, raises costs, and often leads to downcycling or disposal. However, the improper disposal of household waste in Europe is an ongoing issue, causing a massive amount of additional treatment costs and a waste of resources [20] . The overall waste generated by households in the European Union is increasing year by year [5] . In Germany alone, the improper waste disposal in the segment of plastics is up to 30 percent, which corresponds to approximately 750,000 tonnes per year [6] . In Europe, and particularly in Germany, the practical implementation of a high-quality recycling workflow is constrained by complex and heterogeneous local regulat

FAQs

Why is waste sorting difficult for residents in Germany and across Europe?

Waste-disposal regulations and sorting schemes differ across countries and even across German municipalities. This complexity makes correct disposal challenging and reduces the effectiveness of general citizen information systems.

What classification strategies are compared in the study?

The study compares One-vs-All (OvA) and One-vs-Rest (OvR) classification strategies for multi-class waste sorting using deep learning.

How does the confidence-guided human-in-the-loop approach work?

The evaluation applies varying confidence thresholds to select uncertain samples likely to be misclassified. These selected cases are intended for subsequent human review, trading off misclassifications against human annotation effort.

Sep 21, 2026

Uploaded by Ivy

Efficient Waste Sorting for Circular Economy - A Confidence-guided comparison between One-Vs-All and One-Vs-Rest Classification Strategies with Human-in-the-Loop for Automated Waste Sorting_Page 1

Open in WPS to view all 8 pages