Suitability of KANs for Computer Vision: A preliminary investigation
This is an incremental study assessing KANs for computer vision, potentially benefiting researchers in neural network design.
This work investigates the applicability of Kolmogorov-Arnold Networks (KANs) to computer vision tasks like recognition and segmentation, finding that KAN-based architectures perform as claimed but may require more complex edge functions to maintain advantages on complex visual data.
Kolmogorov-Arnold Networks (KANs) introduce a paradigm of neural modeling that implements learnable functions on the edges of the networks, diverging from the traditional node-centric activations in neural networks. This work assesses the applicability and efficacy of KANs in visual modeling, focusing on fundamental recognition and segmentation tasks. We mainly analyze the performance and efficiency of different network architectures built using KAN concepts along with conventional building blocks of convolutional and linear layers, enabling a comparative analysis with the conventional models. Our findings are aimed at contributing to understanding the potential of KANs in computer vision, highlighting both their strengths and areas for further research. Our evaluation point toward the fact that while KAN-based architectures perform in line with the original claims, it may often be important to employ more complex functions on the network edges to retain the performance advantage of KANs on more complex visual data.