Bridging Synthetic and Real-World Domains: A Human-in-the-Loop Weakly-Supervised Framework for Industrial Toxic Emission Segmentation

TL;DR

This paper presents CEDANet, a human-in-the-loop weakly-supervised framework that effectively segments industrial smoke by combining citizen feedback with domain adaptation, significantly reducing the need for costly pixel-level annotations.

Contribution

Introduces CEDANet, a novel framework integrating citizen-provided labels and adversarial domain adaptation for industrial smoke segmentation with minimal supervision.

Findings

Achieves five-fold increase in F1-score over baseline.

Attains performance comparable to limited fully supervised training.

Demonstrates scalability and cost-efficiency in environmental monitoring.

Abstract

Industrial smoke segmentation is critical for air-quality monitoring and environmental protection but is often hampered by the high cost and scarcity of pixel-level annotations in real-world settings. We introduce CEDANet, a human-in-the-loop, class-aware domain adaptation framework that uniquely integrates weak, citizen-provided video-level labels with adversarial feature alignment. Specifically, we refine pseudo-labels generated by a source-trained segmentation model using citizen votes, and employ class-specific domain discriminators to transfer rich source-domain representations to the industrial domain. Comprehensive experiments on SMOKE5K and custom IJmond datasets demonstrate that CEDANet achieves an F1-score of 0.414 and a smoke-class IoU of 0.261 with citizen feedback, vastly outperforming the baseline model, which scored 0.083 and 0.043 respectively. This represents a five-fold increase in F1-score and a six-fold increase in smoke-class IoU. Notably, CEDANet with citizen-constrained pseudo-labels achieves performance comparable to the same architecture trained on limited 100 fully annotated images with F1-score of 0.418 and IoU of 0.264, demonstrating its ability to reach small-sampled fully supervised-level accuracy without target-domain annotations. Our research validates the scalability and cost-efficiency of combining citizen science with weakly supervised domain adaptation, offering a practical solution for complex, data-scarce environmental monitoring applications.