End-to-End Implicit Neural Representations for Classification

Journal Article (2025)
Author(s)

Alexander Gielisse (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Jan Van Gemert (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Research Group
Pattern Recognition and Bioinformatics
DOI related publication
https://doi.org/10.1109/CVPR52734.2025.01745 Final published version
More Info
expand_more
Publication Year
2025
Language
English
Research Group
Pattern Recognition and Bioinformatics
Journal title
Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
Pages (from-to)
18728-18737
Event
2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025 (2025-06-11 - 2025-06-12), Nashville, United States
Page Views
51
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Implicit neural representations (INRs) such as NeRF and SIREN encode a signal in neural network parameters and show excellent results for signal reconstruction. Using INRs for downstream tasks, such as classification, is however not straightforward. Inherent symmetries in the parameters pose challenges and current works primarily focus on designing architectures that are equivariant to these symmetries. However, INR-based classification still significantly under-performs compared to pixel-based methods like CNNs. This work presents an end-to-end strategy for initializing SIRENs together with a learned learning-rate scheme, to yield representations that improve classification accuracy. We show that a simple, straightforward, Transformer model applied to a meta-learned SIREN, without incorporating explicit symmetry equivariances, outperforms the current state-of-the-art. On the CIFAR-10 SIREN classification task, we improve the state-of-the-art without augmentations from 38.8% to 59.6%, and from 63.4% to 64.7% with augmentations. We demonstrate scalability on the high-resolution Imagenette dataset achieving reasonable reconstruction quality with a classification accuracy of 60.8% and are the first to do INR classification on the full ImageNet-1K dataset where we achieve a SIREN classification performance of 23.6%. To the best of our knowledge, no other SIREN classification approach has managed to set a classification baseline for any high-resolution image dataset. Our code is available at https://github.com/SanderGielisse/MWT.

Files

End-to-End_Implicit_Neural_Rep... (pdf)
(pdf | 1.57 Mb)
- Embargo expired in 13-02-2026
– Personal use only – Dutch Copyright Act (Article 25fa)