KARMA: Efficient Structural Defect Segmentation via Kolmogorov-Arnold Representation Learning

2025-08-11

AI Summary

Addressing challenges in semantic segmentation of civil infrastructure defects—including high visual variability, poor imaging conditions, severe class imbalance, and constraints on real-time deployment—this paper proposes TiKAN, a lightweight segmentation architecture. Methodologically, TiKAN introduces three key innovations: (1) a TiKAN module grounded in Kolmogorov–Arnold representation learning, which replaces standard convolutions with compositions of one-dimensional functions; (2) an optimized separable convolution-based feature pyramid; and (3) a static-dynamic prototype mechanism to strengthen representation learning for minority classes. Evaluated on benchmark datasets, TiKAN achieves competitive or superior mIoU while maintaining only 0.959M parameters (97% fewer than state-of-the-art models) and 0.264 GFLOPs computational cost—enabling efficient, real-time inference on edge devices.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Learning on the Edge & Model Compression
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

Semantic segmentation of structural defects in civil infrastructure remains challenging due to variable defect appearances, harsh imaging conditions, and significant class imbalance. Current deep learning methods, despite their effectiveness, typically require millions of parameters, rendering them impractical for real-time inspection systems. We introduce KARMA (Kolmogorov-Arnold Representation Mapping Architecture), a highly efficient semantic segmentation framework that models complex defect patterns through compositions of one-dimensional functions rather than conventional convolutions. KARMA features three technical innovations: (1) a parameter-efficient Tiny Kolmogorov-Arnold Network (TiKAN) module leveraging low-rank factorization for KAN-based feature transformation; (2) an optimized feature pyramid structure with separable convolutions for multi-scale defect analysis; and (3) a static-dynamic prototype mechanism that enhances feature representation for imbalanced classes. Extensive experiments on benchmark infrastructure inspection datasets demonstrate that KARMA achieves competitive or superior mean IoU performance compared to state-of-the-art approaches, while using significantly fewer parameters (0.959M vs. 31.04M, a 97% reduction). Operating at 0.264 GFLOPS, KARMA maintains inference speeds suitable for real-time deployment, enabling practical automated infrastructure inspection systems without compromising accuracy. The source code can be accessed at the following URL: https://github.com/faeyelab/karma.

Problem

• Efficient segmentation of structural defects in infrastructure
• Reducing parameters for real-time inspection systems
• Handling class imbalance in defect appearance segmentation

Innovation

• Tiny Kolmogorov-Arnold Network for efficient feature transformation
• Optimized feature pyramid with separable convolutions
• Static-dynamic prototype mechanism for imbalanced classes