WasteAssistant: Regulation-Guided Visual Question Answering Framework for Intelligent Waste Segregation and Sustainable Managemen

2026-07-12

AI Summary

This study addresses the limitations of existing automated waste sorting systems, which often rely solely on unimodal visual inputs, lack contextual understanding, and struggle to align with regulatory requirements, thereby hindering efficient source separation. To overcome these challenges, this work proposes a visual question answering framework that integrates vision-language models with multimodal large language models, uniquely embedding India’s Solid Waste Management Rules (2016) into the multimodal reasoning process to enable regulation-aware joint visual-linguistic understanding. The contributions include an extensible AI solution tailored for urban waste management, the release of WasteVQA—a novel dataset comprising 21 waste categories—and state-of-the-art performance on this benchmark, achieving BLEU 0.8291 and BERTScore 0.9273, substantially outperforming conventional CNN-based approaches and enhancing both classification accuracy and municipal deployment feasibility.

Abstract

Efficient waste segregation is critical for sustainable urban management and environmental governance. Existing automated systems are limited by single-modality visual processing, insufficient contextual understanding, and weak regulatory alignment. To address these issues, we propose a language-guided vision-AI framework that integrates vision-language models and multimodal large language models for joint visual-linguistic reasoning. This framework implements a visual question answering paradigm aligned with India's Solid Waste Management Rules 2016. We construct a new WasteVQA dataset with 13,500 question-answer pairs across 21 waste categories. Experiments show that the BLIP-based model achieves a BLEU score of 0.8291 and a BERTScore of 0.9273, outperforming traditional CNN-based methods. This work improves source-level segregation accuracy, ensures regulatory compliance, and supports scalable deployment for municipal and citizen-facing waste management, promoting multimodal AI in sustainable urban infrastructure. The source code and dataset are available at: https://github.com/Khushkataruka/WasteAssistant

Problem

• waste segregation
• visual question answering
• regulatory compliance
• multimodal AI
• sustainable urban management

Innovation

• Regulation-Guided AI
• Multimodal Visual Question Answering
• Vision-Language Models
• Waste Segregation
• Sustainable Urban Management