What question did this study set out to answer?

To enhance human pose estimation accuracy in coal mining using a self-learning image-conditioned network.

March 13, 2026Journal of King Saud University - Computer and Information Sciences0 citationsOpen Access

SLIC-POSE: Self-learning image-conditioned human pose estimation network for real-time coal mining safety monitoring

Key Points

To enhance human pose estimation accuracy in coal mining using a self-learning image-conditioned network.
Developed a Differentiable Self-learning Image Restoration Module (DSIRM) for image enhancement.
Employed Context-Aware Parameter Learner and Differentiable Restoration Layer for preprocessing.
Introduced Channel Importance-aware Knowledge Distillation (CIKD) for model compression and selective knowledge transfer.
Conducted experiments to assess performance across YOLO and OKS metrics.
Achieved robust pose estimation performance despite challenging mining conditions.
Demonstrated 67 FPS capability on edge devices in real-time applications.
Successfully deployed in six diverse mining scenarios for hazard monitoring.

Abstract

Human pose estimation in coal mining environments faces challenges from pervasive coal dust interference and non-uniform illumination conditions, which create composite noise patterns that degrade pose estimation accuracy. Although existing methods have improved robustness through deep learning architectures, their substantial computational demands hinder practical deployment in mining scenarios. To address these challenges, we propose SLIC-POSE, a novel Self-Learning Image-Conditioned human pose estimation network specifically designed for coal mining environments. First, we design a Differentiable Self-learning Image Restoration Module (DSIRM) that operates as a learnable preprocessing stage, dynamically rectifying photometric and structural degradation without requiring paired training data. DSIRM employs a Context-Aware Parameter Learner and a Differentiable Restoration Layer to enhance images. It sequentially performs self-adaptive de-dusting, self-calibrating illumination, and feature re-projection. These processes are optimized end-to-end using feedback from downstream tasks. Second, we present a Channel Importance-aware Knowledge Distillation (CIKD) framework that systematically evaluates channel significance through group-wise L1-based metrics to guide selective knowledge transfer. By constructing adaptive knowledge filters that emphasize high-value channels while attenuating low-contribution features, CIKD achieves efficient model compression without sacrificing essential representation capabilities. Experimental results demonstrate that SLIC-POSE delivers robust performance across both YOLO and OKS metrics. Furthermore, we demonstrate the practical effectiveness of SLIC-POSE via FlexZone, a real-time hazardous area intrusion warning system. This system has been deployed across six diverse mining scenarios and achieves 67 FPS on edge devices, enabling proactive safety management in coal mining.

问 AI

Bookmark

View Full Paper