Enhanced Seam Segmentation for Automated Welding Robot in Construction Through Transfer Learning: Addressing Limitations of Bilateral Segmentation Network

2026-07-07

AI Summary

This study addresses the significant performance degradation in thin weld seam segmentation within architectural scenes under strong illumination and specular reflection. Building upon the lightweight BiSeNetV2 architecture, the authors propose an optimization strategy that requires no structural modification to the model, integrating transfer learning, a hybrid cross-entropy–Lovász loss, and OHEM-based contrastive refinement. This approach substantially enhances robustness to reflective environments while preserving real-time inference speed. Experimental results demonstrate that the method achieves 81.76% Joint IoU and 90.73% mIoU on the test set, representing a 22.36-percentage-point improvement in Joint IoU over the baseline and successfully recovering 96.33% of previously failed samples with zero IoU.

Abstract

Reliable seam segmentation is essential for autonomous robotic welding in construction, where harsh illumination, specular reflections, and thin weld geometries often degrade segmentation performance. This study proposes a reflection-robust seam segmentation framework that enhances a BiSeNetV2 backbone through transfer learning and a hybrid Cross-Entropy--Lovász loss. Rather than increasing architectural complexity, the proposed framework improves reflection robustness through learning-stability-oriented optimization. Experimental results show that the proposed method achieves 81.76\% Joint IoU and 90.73\% mIoU, improving Joint IoU by +22.36 percentage points over the OHEM-based baseline while maintaining identical FLOPs, parameter count, and inference speed. The proposed approach also recovers 96.33\% of severe zero-IoU failure cases under reflective conditions. Comparative experiments across BiSeNetV2, DeepLabV3+, UNet, and SegFormer further demonstrate that the proposed optimization strategy is particularly effective for lightweight real-time segmentation architectures. Qualitative analyses additionally show improved seam continuity and reflection robustness in challenging welding environments. These findings suggest that the proposed framework provides a practical and lightweight perception solution for robotic welding applications involving reflective metallic surfaces.

Problem

• seam segmentation
• welding robot
• specular reflections
• construction automation
• robust perception

Innovation

• transfer learning
• seam segmentation
• reflection robustness
• lightweight architecture
• hybrid loss function