Taylor expansion-based Kolmogorov-Arnold network for blind image quality assessment

2025-05-27

AI Summary

To address the limited performance gains and high computational overhead of Kolmogorov–Arnold Networks (KANs) in blind image quality assessment (BIQA) when processing high-dimensional features, this paper proposes TaylorKAN. It introduces learnable Taylor series expansions as local activation functions within the KAN architecture—marking the first such integration—and thereby overcomes the limitations of global approximation imposed by conventional orthogonal basis functions. Furthermore, we design a synergistic optimization framework combining depth reduction and dimensionality compression to achieve model lightweighting and efficient feature modeling. Evaluated on five real-world distorted image datasets—BID, CLIVE, KonIQ, SPAQ, and FLIVE—TaylorKAN consistently outperforms existing KAN variants in score regression tasks, demonstrates significantly improved cross-dataset generalization, and achieves higher inference efficiency.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Machine Learning: Deep Neural Architectures and Foundation Models
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Kolmogorov-Arnold Network (KAN) has attracted growing interest for its strong function approximation capability. In our previous work, KAN and its variants were explored in score regression for blind image quality assessment (BIQA). However, these models encounter challenges when processing high-dimensional features, leading to limited performance gains and increased computational cost. To address these issues, we propose TaylorKAN that leverages the Taylor expansions as learnable activation functions to enhance local approximation capability. To improve the computational efficiency, network depth reduction and feature dimensionality compression are integrated into the TaylorKAN-based score regression pipeline. On five databases (BID, CLIVE, KonIQ, SPAQ, and FLIVE) with authentic distortions, extensive experiments demonstrate that TaylorKAN consistently outperforms the other KAN-related models, indicating that the local approximation via Taylor expansions is more effective than global approximation using orthogonal functions. Its generalization capacity is validated through inter-database experiments. The findings highlight the potential of TaylorKAN as an efficient and robust model for high-dimensional score regression.

Problem

• Enhancing KAN's local approximation for blind image quality assessment
• Reducing computational cost in high-dimensional feature processing
• Improving generalization in score regression for BIQA

Innovation

• Taylor expansions as learnable activation functions
• Network depth reduction for efficiency
• Feature dimensionality compression integration