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

State Space Representation01:27

State Space Representation

163
The frequency-domain technique, commonly used in analyzing and designing feedback control systems, is effective for linear, time-invariant systems. However, it falls short when dealing with nonlinear, time-varying, and multiple-input multiple-output systems. The time-domain or state-space approach addresses these limitations by utilizing state variables to construct simultaneous, first-order differential equations, known as state equations, for an nth-order system.
Consider an RLC circuit, a...
163
Self-Discrepancy Theory02:45

Self-Discrepancy Theory

18.3K
One influential perspective on what motivates people's behavior is detailed in Tory Higgin's self-discrepancy theory (Higgins, 1987). He proposed that people hold disagreeing internal representations of themselves that lead to different emotional states.  
18.3K
Difference from Background: Limit of Detection01:05

Difference from Background: Limit of Detection

5.9K
The limit of detection (LOD) is the smallest amount of analyte that can be distinguished from the background noise. The LOD value corresponds to the concentration at which the analyte signal is three times larger than the standard deviation of the blank signal. Below this value, the analyte signal cannot be differentiated from the background noise. It is calculated by dividing the calibration slope by 3 times the standard deviation of the blank signals.
The LOD indicates the presence or absence...
5.9K
The Representativeness Heuristic02:13

The Representativeness Heuristic

15.8K
The representative heuristic describes a biased way of thinking, in which you unintentionally stereotype someone or something. For example, you may assume that your professors spend their free time reading books and engaging in intellectual conversation, because the idea of them spending their time playing volleyball or visiting an amusement park does not fit in with your stereotypes of professors.
15.8K
Nonconscious Mimicry01:13

Nonconscious Mimicry

4.5K
Nonconscious mimicry occurs when individuals alter their mannerisms to match the behaviors and expressions of those nearby, without intention.
4.5K
Upsampling01:22

Upsampling

203
Managing signal sampling rates is essential in digital signal processing to maintain signal integrity. A decimated signal, characterized by a reduced frequency range due to its lower sampling rate, can be upsampled by inserting zeros between each sample. This upsampling process expands the original spectrum and introduces repeated spectral replicas at intervals dictated by the new Nyquist frequency. To refine this zero-inserted sequence, it is passed through a lowpass filter with a cutoff...
203

You might also read

Related Articles

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

Sort by
Same author

Biological direct-shortcut deep residual learning for sparse visual processing.

Scientific reports·2024
See all related articles

Related Experiment Video

Updated: Jun 5, 2025

Defining the Role Of Language in Infants' Object Categorization with Eye-tracking Paradigms
07:31

Defining the Role Of Language in Infants' Object Categorization with Eye-tracking Paradigms

Published on: February 8, 2019

6.5K

Kalman contrastive unsupervised representation learning.

Mohammad Mahdi Jahani Yekta1

  • 1Department of Computer Science, Stanford University, 353 Jane Stanford Way, Stanford, CA, 94305, USA. m_mahdi_jahani@yahoo.com.

Scientific Reports
|December 5, 2024
PubMed
Summary

Kalman contrastive (KalCo) framework enhances unsupervised representation learning using a dynamic dictionary. KalCo significantly outperforms momentum contrastive (MoCo) learning, achieving higher accuracy on various datasets.

Keywords:
Contrastive unsupervised learningDictionary buildingKalman filterMoCoRegularized optimization

More Related Videos

A Psychophysics Paradigm for the Collection and Analysis of Similarity Judgments
08:12

A Psychophysics Paradigm for the Collection and Analysis of Similarity Judgments

Published on: March 1, 2022

2.4K
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

468

Related Experiment Videos

Last Updated: Jun 5, 2025

Defining the Role Of Language in Infants' Object Categorization with Eye-tracking Paradigms
07:31

Defining the Role Of Language in Infants' Object Categorization with Eye-tracking Paradigms

Published on: February 8, 2019

6.5K
A Psychophysics Paradigm for the Collection and Analysis of Similarity Judgments
08:12

A Psychophysics Paradigm for the Collection and Analysis of Similarity Judgments

Published on: March 1, 2022

2.4K
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

468

Area of Science:

  • Computer Science
  • Machine Learning
  • Artificial Intelligence

Background:

  • Unsupervised representation learning is crucial for leveraging large unlabeled datasets.
  • Existing methods like momentum contrastive (MoCo) learning have limitations in accuracy and consistency.
  • Dynamic dictionary learning offers a promising avenue for improving representation quality.

Purpose of the Study:

  • To introduce a novel Kalman contrastive (KalCo) framework for unsupervised representation learning.
  • To enhance representation learning accuracy by employing a dynamic dictionary and Kalman filter.
  • To compare KalCo's performance against established methods like MoCo.

Main Methods:

  • Developed the Kalman contrastive (KalCo) framework utilizing a dynamic dictionary with a queue and Kalman filter encoder.
  • Implemented KalCo for unsupervised representation learning on instance discrimination pretext tasks.
  • Upgraded the framework to KalCo v2, incorporating an MLP projection head, enhanced data augmentation, and a larger memory bank.

Main Results:

  • KalCo achieved 80% accuracy on ImageNet-1M (IN-1M), significantly surpassing MoCo's 55%.
  • Comparable high accuracy was observed on Instagram-1B (IG-1B) and OpenfMRI datasets (84%).
  • KalCo v2 reached 90% accuracy on IN-1M and IG-1B, and 95% on OpenfMRI, outperforming recent alternatives.

Conclusions:

  • The Kalman contrastive (KalCo) framework provides a robust and accurate approach to unsupervised representation learning.
  • KalCo's dynamic dictionary mechanism is key to its superior performance over methods like MoCo.
  • KalCo v2 represents a significant advancement, setting new benchmarks in unsupervised learning accuracy.