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 Concept Videos

Survival Tree01:19

Survival Tree

133
Survival trees are a non-parametric method used in survival analysis to model the relationship between a set of covariates and the time until an event of interest occurs, often referred to as the "time-to-event" or "survival time." This method is particularly useful when dealing with censored data, where the event has not occurred for some individuals by the end of the study period, or when the exact time of the event is unknown.
 Building a Survival Tree
Constructing a...
133
Improving Translational Accuracy02:07

Improving Translational Accuracy

2.7K
2.7K
Reducing Line Loss01:18

Reducing Line Loss

188
In a three-phase circuit, line loss is an indicator of energy dissipated as heat due to the resistance of transmission lines. To address this, incorporating transformers into the system—a step-up transformer at the source and a step-down transformer at the load—is a strategic solution. Two three-phase transformers are introduced to improve this.
With a step-up transformer at the source, the voltage is increased, thereby reducing the current in the transmission lines since power loss...
188
Gradient and Del Operator01:14

Gradient and Del Operator

2.8K
In mathematics and physics, the gradient and del operator are fundamental concepts used to describe the behavior of functions and fields in space. The gradient is a mathematical operator that gives both the magnitude and direction of the maximum spatial rate of change. Consider a person standing on a mountain. The slope of the mountain at any given point is not defined unless it is quantified in a particular direction. For this reason, a "directional derivative" is defined, which is a vector...
2.8K
Observational Learning01:12

Observational Learning

280
Albert Bandura's observational learning, also known as imitation or modeling, occurs when a person observes and imitates another's behavior. It is a quicker process than operant conditioning. A well-known example is the Bobo doll study, where children who saw an adult acting aggressively towards the doll were more likely to act aggressively when left alone, compared to those who observed a nonaggressive adult. Many psychologists view observational learning as a form of latent learning...
280

You might also read

Related Articles

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

Sort by
Same author

A feature optimization study based on a diabetes risk questionnaire.

Frontiers in public health·2024
Same author

Diabetes prediction model for unbalanced community follow-up data set based on optimal feature selection and scorecard.

Digital health·2024
Same author

Diabetes risk prediction model based on community follow-up data using machine learning.

Preventive medicine reports·2023
Same author

Broadband high quantum efficiency InGaAs/InP focal plane arrays via high precision plasma thinning.

Optics letters·2020
Same author

[TOTAL KNEE ARTHROPLASTY IN YOUNG PATIENTS WITH OSTEOARTHRITIS].

Zhongguo xiu fu chong jian wai ke za zhi = Zhongguo xiufu chongjian waike zazhi = Chinese journal of reparative and reconstructive surgery·2015
Same author

Interferon-γ and Tumor Necrosis Factor-α Produced by T Cells Reduce the HBV Persistence Form, cccDNA, Without Cytolysis.

Gastroenterology·2015

Related Experiment Video

Updated: Aug 25, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.3K

Recover User's Private Training Image Data by Gradient in Federated Learning.

Haimei Gong1,2, Liangjun Jiang1, Xiaoyang Liu1

  • 1College of Information and Communication Engineering, Hainan University, Haikou 570228, China.

Sensors (Basel, Switzerland)
|October 14, 2022
PubMed
Summary

Gradient inversion attacks pose a privacy risk in machine learning. This study introduces a new system, SSRAS, that reconstructs private training images with high fidelity, even from convolutional neural networks.

Keywords:
Federated Learningdata reconstruction attackgradient leakage attackssecurity and privacy

More Related Videos

Author Spotlight: Efficient Image Recognition Using Directional Gradient Histogram Technique and Support Vector Machines
08:27

Author Spotlight: Efficient Image Recognition Using Directional Gradient Histogram Technique and Support Vector Machines

Published on: January 5, 2024

1.2K

Related Experiment Videos

Last Updated: Aug 25, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.3K
Author Spotlight: Efficient Image Recognition Using Directional Gradient Histogram Technique and Support Vector Machines
08:27

Author Spotlight: Efficient Image Recognition Using Directional Gradient Histogram Technique and Support Vector Machines

Published on: January 5, 2024

1.2K

Area of Science:

  • Machine Learning Security
  • Data Privacy
  • Computer Vision

Background:

  • Gradient exchange is common in distributed machine learning systems.
  • Gradients were previously assumed safe for transmission.
  • Recent research indicates gradient inversion can reconstruct input data.

Purpose of the Study:

  • To review and categorize existing gradient inversion attacks.
  • To propose a novel privacy attack system, SSRAS.
  • To develop a metric (RA-I) for assessing reconstruction vulnerability.

Main Methods:

  • Categorization of gradient inversion techniques into Bias Attacks, Optimization-Based Attacks, and Linear Equation Solver Attacks.
  • Development of the Single-Sample Reconstruction Attack System (SSRAS).
  • Proposal of the Improved R-GAP Algorithm and Rank Analysis Index (RA-I).

Main Results:

  • SSRAS successfully reconstructs private training images with high fidelity.
  • The system extends gradient inversion to various network layers and types (fully connected, CNNs) with or without bias.
  • RA-I effectively measures the potential for raw image data reconstruction.

Conclusions:

  • Gradient inversion attacks represent a significant privacy threat in machine learning.
  • SSRAS demonstrates superior performance in reconstructing private data compared to existing methods.
  • The proposed methods enhance the understanding and quantification of privacy risks in gradient-based systems.