FM-SIREN & FM-FINER: Implicit Neural Representation Using Nyquist-based Orthogonality

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Alsakabi, Mohammed, Mobeirek, Wael, Dolan, John M., Tonguz, Ozan K.
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913023321440256
author Alsakabi, Mohammed
Mobeirek, Wael
Dolan, John M.
Tonguz, Ozan K.
author_facet Alsakabi, Mohammed
Mobeirek, Wael
Dolan, John M.
Tonguz, Ozan K.
contents Existing periodic activation-based implicit neural representation (INR) networks, such as SIREN and FINER, suffer from hidden feature redundancy, where neurons within a layer capture overlapping frequency components due to the use of a fixed frequency multiplier. This redundancy limits the expressive capacity of multilayer perceptrons (MLPs). Drawing inspiration from classical signal processing methods such as the Discrete Sine Transform (DST), in this paper, we propose FM-SIREN and FM-FINER, which assign Nyquist-informed, neuron-specific frequency multipliers to periodic activations. Contrary to existing approaches, our design introduces frequency diversity without requiring hyperparameter tuning or additional network depth. This simple yet principled approach reduces the redundancy of features by nearly 50% and consistently improves signal reconstruction across diverse INR tasks, such as fitting 1D audio, 2D image and 3D shape, and video, outperforming their baseline counterparts while maintaining efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2509_23438
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FM-SIREN & FM-FINER: Implicit Neural Representation Using Nyquist-based Orthogonality
Alsakabi, Mohammed
Mobeirek, Wael
Dolan, John M.
Tonguz, Ozan K.
Computer Vision and Pattern Recognition
Existing periodic activation-based implicit neural representation (INR) networks, such as SIREN and FINER, suffer from hidden feature redundancy, where neurons within a layer capture overlapping frequency components due to the use of a fixed frequency multiplier. This redundancy limits the expressive capacity of multilayer perceptrons (MLPs). Drawing inspiration from classical signal processing methods such as the Discrete Sine Transform (DST), in this paper, we propose FM-SIREN and FM-FINER, which assign Nyquist-informed, neuron-specific frequency multipliers to periodic activations. Contrary to existing approaches, our design introduces frequency diversity without requiring hyperparameter tuning or additional network depth. This simple yet principled approach reduces the redundancy of features by nearly 50% and consistently improves signal reconstruction across diverse INR tasks, such as fitting 1D audio, 2D image and 3D shape, and video, outperforming their baseline counterparts while maintaining efficiency.
title FM-SIREN & FM-FINER: Implicit Neural Representation Using Nyquist-based Orthogonality
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2509.23438