Polynomial Neural Fields for Subband Decomposition and Manipulation
arXiv:2302.04862
Abstract
Neural fields have emerged as a new paradigm for representing signals, thanks to their ability to do it compactly while being easy to optimize. In most applications, however, neural fields are treated like black boxes, which precludes many signal manipulation tasks. In this paper, we propose a new class of neural fields called polynomial neural fields (PNFs). The key advantage of a PNF is that it can represent a signal as a composition of a number of manipulable and interpretable components without losing the merits of neural fields representation. We develop a general theoretical framework to analyze and design PNFs. We use this framework to design Fourier PNFs, which match state-of-the-art performance in signal representation tasks that use neural fields. In addition, we empirically demonstrate that Fourier PNFs enable signal manipulation applications such as texture transfer and scale-space interpolation. Code is available at https://github.com/stevenygd/PNF.
Accepted to NeurIPS 2022
References in corpus (13)
- Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
- Squeeze-and-Excitation Networks
- Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains
- NeRF++: Analyzing and Improving Neural Radiance Fields
- DONeRF: Towards Real-Time Rendering of Compact Neural Radiance Fields using Depth Oracle Networks
- On the Effectiveness of Weight-Encoded Neural Implicit 3D Shapes
- On the Expressive Power of Deep Polynomial Neural Networks
- Rethinking Positional Encoding
- PolyGAN: High-Order Polynomial Generators
- Survey of Expressivity in Deep Neural Networks
- MINER: Multiscale Implicit Neural Representations
- Multilinear Latent Conditioning for Generating Unseen Attribute Combinations
- Adversarial Audio Synthesis with Complex-valued Polynomial Networks