AI Article Synopsis
- Visual neurons and visual perception are complex and often nonlinear, contradicting most current vision models that use a simpler linear receptive field (RF).
- The traditional linear RF has limitations, like changing with different inputs and not aligning with recent findings about how neurons process information.
- The proposed intrinsically nonlinear receptive field (INRF) offers a more accurate model, maintaining consistency across various stimuli, leading to better performance in artificial neural networks and suggesting a shift in both vision science and artificial intelligence approaches.
Article Abstract
The responses of visual neurons, as well as visual perception phenomena in general, are highly nonlinear functions of the visual input, while most vision models are grounded on the notion of a linear receptive field (RF). The linear RF has a number of inherent problems: it changes with the input, it presupposes a set of basis functions for the visual system, and it conflicts with recent studies on dendritic computations. Here we propose to model the RF in a nonlinear manner, introducing the intrinsically nonlinear receptive field (INRF). Apart from being more physiologically plausible and embodying the efficient representation principle, the INRF has a key property of wide-ranging implications: for several vision science phenomena where a linear RF must vary with the input in order to predict responses, the INRF can remain constant under different stimuli. We also prove that Artificial Neural Networks with INRF modules instead of linear filters have a remarkably improved performance and better emulate basic human perception. Our results suggest a change of paradigm for vision science as well as for artificial intelligence.
Download full-text PDF |
Source |
|---|---|
| http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7530701 | PMC |
| http://dx.doi.org/10.1038/s41598-020-73113-0 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Preparation of ZnAl layered double hydroxides supported by silica for the treatment of Cr(VI) and Cu(II) in aqueous solution.
Sci Rep
January 2025
School of Chemistry and Chemical Engineering, Heze University, Heze, 274015, People's Republic of China.
A novel adsorbent ZnAl-LDHs/SiO (ZA/SiO) was prepared by blending urea mixture of ZnSO and Al(SO) while using SiO as a support form. The adsorption properties of ZA/SiO for the removal of toxic metal ions (Cu(II) and Cr(VI)) from water were evaluated. By batch experiment method to investigate the ZA/SiO adsorption of Cu(II) and Cr(VI) solution treatment effect.
View Article and Find Full Text PDFFerroelectricity and Nonlinear Optical Responses in Two-Dimensional Distorted ( = Cu, Ag, Au; = Chalcogens; = Halogen) Monolayers.
ACS Appl Mater Interfaces
January 2025
Department of Physics, School of Physical Science and Engineering, Beijing Jiaotong University, Beijing 100044, China.
Two-dimensional (2D) materials with spontaneous polarization can exhibit large second-order nonlinear optical (NLO) effects. Here, we present a series of stable distorted monolayers by using first-principles calculations and lattice vibration analysis. The structural distortion leads to a lower polar symmetry, giving rise to intrinsic ferroelectricity with a Curie point up to room temperature.
View Article and Find Full Text PDFThe development of diffusion models, such as Glide, DALLE 2, Imagen, and Stable Diffusion, marks a significant advancement in generative AI for image synthesis. In this paper, we introduce a novel framework for synthesizing intrinsic connectivity networks (ICNs) by utilizing the nonlinear capabilities of denoising diffusion probabilistic models (DDPMs). This approach builds upon and extends traditional linear methods, such as independent component analysis (ICA), which are commonly used in neuroimaging studies.
View Article and Find Full Text PDFNonlinear feature selection for support vector quantile regression.
Neural Netw
January 2025
School of Economics, Zhejiang University of Technology, Hangzhou 310023, China; Institute for Industrial System Modernization, Zhejiang University of Technology, Hangzhou 310023, China; Zhijiang College, Zhejiang University of Technology, Shaoxing 312030, China. Electronic address:
This paper discusses the nuanced domain of nonlinear feature selection in heterogeneous systems. To address this challenge, we present a sparsity-driven methodology, namely nonlinear feature selection for support vector quantile regression (NFS-SVQR). This method includes a binary-diagonal matrix, featuring 0 and 1 elements, to address the complexities of feature selection within intricate nonlinear systems.
View Article and Find Full Text PDFEfficient Second Harmonic Generation via Plasmonic-Photonic Mode Matching in Hybrid Waveguide.
Nano Lett
January 2025
School of Physics and Technology, and Key Laboratory of Artificial Micro- and Nano-structures of Ministry of Education, Wuhan University, Wuhan 430072, China.
Hybrid nonlinear plasmonic waveguides, characterized by a small mode area and large nonlinear susceptibility, present an intriguing and practical platform for the minimization of nonlinear photonic devices. Nevertheless, the intrinsic Ohmic loss associated with surface plasmon polaritons (SPPs) and modal dispersion imposes constraints on the effective interaction length and, consequently, the ultimate efficiency of nonlinear processes. In this study, we demonstrate an efficient second harmonic generation (SHG) within a hybrid plasmonic waveguide by leveraging SPP-like modes at the fundamental wave and photonic-like modes at the SHG under phase matching conditions.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!