Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jan 10, 2026

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

999

Multilayer nonlinear diffraction neural networks with programmable and fast ReLU activation function.

Yu Ming Ning1,2,3, Qian Ma4,5, Qiang Xiao1,2

  • 1State Key Laboratory of Millimeter Waves and Institute of Electromagnetic Space, Southeast University, Nanjing, China.

Nature Communications
|November 24, 2025
PubMed
Summary

Related Concept Videos

Neural Circuits01:25

Neural Circuits

2.6K
Neural circuits and neuronal pools are two of the main structures found in the nervous system. Neural circuits are networks of neurons that work together to carry out a specific task or process. They consist of interconnected neurons and glial cells, which provide structural and metabolic support.
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
2.6K

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Generative model for information metamaterial design.

Nature computational science·2026
Same author

Serum lipidome remodeling in viral pneumonia: from pathophysiology to therapeutics.

Frontiers in immunology·2026
Same author

Rhizome differentiation is associated with metabolic specialization and rhizosphere microbial assembly in Rheum officinale Baill.

Plant cell reports·2026
Same author

Space-Time Coding Conformal Metasurfaces for Multifrequency Beam Steering and Shaping.

Advanced materials (Deerfield Beach, Fla.)·2026
Same author

Exploring the metabolic characteristics of fresh-cut sweet potatoes at different storage times based on untargeted metabolomics.

Food chemistry: X·2026
Same author

Conformal, transparent, and efficient MXene grid antennas for flexible wireless electronics.

Science advances·2026

We developed a novel microwave nonlinear neural network using programmable metasurfaces. This innovation significantly boosts speed and reduces power consumption for machine learning applications like image classification.

Area of Science:

  • Artificial Intelligence
  • Electromagnetics
  • Materials Science

Background:

  • Optical diffractive neural networks offer speed and energy efficiency gains.
  • Nonlinear activation functions present challenges like latency and high power consumption, hindering practical deployment.

Purpose of the Study:

  • To propose a programmable multilayer full-space nonlinear neural network operating in the microwave frequency band.
  • To overcome the limitations of existing nonlinear activation functions in neural networks.

Main Methods:

  • Constructed nonlinear layers using programmable metasurfaces integrated with RF components.
  • Implemented a ReLU-like nonlinear activation function.
  • Operated the system in the microwave frequency band.

Related Experiment Videos

Last Updated: Jan 10, 2026

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

999

Main Results:

  • Achieved a nanosecond-scale delay (17.7 ns), significantly faster than photoelectric methods.
  • Demonstrated reconfigurable nonlinear activation functions with low thresholds.
  • Showcased effective image classification and real-time human posture recognition.

Conclusions:

  • The proposed microwave nonlinear neural network offers low latency, high speed, and low power consumption.
  • The flexible nonlinear activation capabilities make it promising for various applications.
  • This architecture advances the practical deployment of efficient machine learning.