Hyb-KAN ViT: Hybrid Kolmogorov-Arnold Networks Augmented Vision Transformer

TL;DR

Hyb-KAN ViT introduces a modular hybrid architecture combining wavelet spectral decomposition and spline activations to improve multi-scale feature extraction and efficiency in vision transformer tasks, achieving state-of-the-art results.

Contribution

The paper presents a novel hybrid ViT framework integrating wavelet transforms and spline functions, enhancing spectral modeling and modularity over prior architectures.

Findings

State-of-the-art performance on ImageNet-1K, COCO, and ADE20K datasets.

Wavelet spectral priors improve segmentation accuracy.

Spline-based modules enhance detection efficiency.

Abstract

This study addresses the inherent limitations of Multi-Layer Perceptrons (MLPs) in Vision Transformers (ViTs) by introducing Hybrid Kolmogorov-Arnold Network (KAN)-ViT (Hyb-KAN ViT), a novel framework that integrates wavelet-based spectral decomposition and spline-optimized activation functions, prior work has failed to focus on the prebuilt modularity of the ViT architecture and integration of edge detection capabilities of Wavelet functions. We propose two key modules: Efficient-KAN (Eff-KAN), which replaces MLP layers with spline functions and Wavelet-KAN (Wav-KAN), leveraging orthogonal wavelet transforms for multi-resolution feature extraction. These modules are systematically integrated in ViT encoder layers and classification heads to enhance spatial-frequency modeling while mitigating computational bottlenecks. Experiments on ImageNet-1K (Image Recognition), COCO (Object Detection and Instance Segmentation), and ADE20K (Semantic Segmentation) demonstrate state-of-the-art performance with Hyb-KAN ViT. Ablation studies validate the efficacy of wavelet-driven spectral priors in segmentation and spline-based efficiency in detection tasks. The framework establishes a new paradigm for balancing parameter efficiency and multi-scale representation in vision architectures.