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

Aggregates Classification01:29

Aggregates Classification

317
Aggregate classification is generally based on its size, petrographic characteristics, weight, and source. Size classification ranges from coarse to fine aggregates, defined by the size of the particles. Coarse aggregates are particles that do not pass through ASTM sieve No. 4, and aggregates that pass through the sieve are fine aggregates.
Petrographic classification groups aggregates based on common mineralogical characteristics. Some of the common mineral groups found in aggregates are...
317
Classification of Signals01:30

Classification of Signals

435
In signal processing, signals are classified based on various characteristics: continuous-time versus discrete-time, periodic versus aperiodic, analog versus digital, and causal versus noncausal. Each category highlights distinct properties crucial for understanding and manipulating signals.
A continuous-time signal holds a value at every instant in time, representing information seamlessly. In contrast, a discrete-time signal holds values only at specific moments, often denoted as x(n), where...
435
Classification of Systems-I01:26

Classification of Systems-I

179
Linearity is a system property characterized by a direct input-output relationship, combining homogeneity and additivity.
Homogeneity dictates that if an input x(t) is multiplied by a constant c, the output y(t) is multiplied by the same constant. Mathematically, this is expressed as:
179
Force Classification01:22

Force Classification

1.2K
Forces play a crucial role in the study of physics and engineering. They are essential in describing the motion, behavior, and equilibrium of objects in the physical world. Forces can be classified based on their origin, type, and direction of action.
Contact and non-contact forces are two of the most widely used categories of forces. As the name suggests, contact forces require physical contact between two objects to act upon each other. Examples of contact forces include frictional,...
1.2K
Classification of Systems-II01:31

Classification of Systems-II

139
Continuous-time systems have continuous input and output signals, with time measured continuously. These systems are generally defined by differential or algebraic equations. For instance, in an RC circuit, the relationship between input and output voltage is expressed through a differential equation derived from Ohm's law and the capacitor relation,
139
Weighted Mean00:57

Weighted Mean

5.1K
While taking the arithmetic, geometric, or harmonic mean of a sample data set, equal importance is assigned to all the data points. However, all the values may not always be equally important in some data sets. An intrinsic bias might make it more important to give more weightage to specific values over others.
For example, consider the number of goals scored in the matches of a tournament. While computing the average number of goals scored in the tournament, it may be more important to...
5.1K

You might also read

Related Articles

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

Sort by
Same author

Modulation of 1D Ru-Porphyrin Biomimetic COF to Enhance Synergistic Dual C─H Bond Air Oxidation for Cyclohexenone Synthesis.

Small (Weinheim an der Bergstrasse, Germany)·2026
Same author

Programmable Control Over Radical and Non‑Radical Pathways in Fenton‑Like Catalysis via Carbon‑Encapsulated Iron Nanoreactors.

Small (Weinheim an der Bergstrasse, Germany)·2026
Same author

Serum syndecan-1 correlates with coronary artery calcification severity and intradialytic hypotension in elderly hemodialysis patients.

Frontiers in cardiovascular medicine·2026
Same author

Synergistic thermodynamic-kinetic tuning of spinel oxides achieve robust seawater hydrogen evolution reaction.

Journal of colloid and interface science·2026
Same author

Highly efficient and stable perovskite photovoltaics enabled by multifunctional crosslinked n<sup>+</sup>-type interlayer.

Nature communications·2026
Same author

Density-Energy Balanced Selection of Chiral Honeycomb Superstructures from Deprotonated Carboxylate Molecules on Ag(111).

ACS applied materials & interfaces·2026

Related Experiment Video

Updated: Jun 22, 2025

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction
06:19

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction

Published on: August 16, 2024

418

Cluster-CAM: Cluster-weighted visual interpretation of CNNs' decision in image classification.

Zhenpeng Feng1, Hongbing Ji1, Miloš Daković2

  • 1School of Electronic Engineering, Xidian University, Xi'an, China.

Neural Networks : the Official Journal of the International Neural Network Society
|June 28, 2024
PubMed
Summary

Cluster-CAM is a novel gradient-free algorithm for interpreting convolutional neural networks (CNNs). This method enhances visualization by clustering feature maps, improving accuracy and efficiency in computer vision tasks.

Keywords:
Class activation mappingClustering algorithmExplainable artificial intelligenceImage classification

More Related Videos

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

515
Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation
08:47

Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation

Published on: February 9, 2024

1.4K

Related Experiment Videos

Last Updated: Jun 22, 2025

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction
06:19

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction

Published on: August 16, 2024

418
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

515
Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation
08:47

Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation

Published on: February 9, 2024

1.4K

Area of Science:

  • Computer Vision
  • Artificial Intelligence
  • Machine Learning

Background:

  • Convolutional Neural Networks (CNNs) are highly successful in computer vision but lack clear interpretability.
  • Class Activation Mapping (CAM) techniques visualize CNN decisions, with gradient-based methods facing issues like vanishing/exploding gradients, and gradient-free methods being time-consuming.

Purpose of the Study:

  • To propose Cluster-CAM, an effective and efficient gradient-free algorithm for CNN interpretation.
  • To address the limitations of existing CAM techniques by improving both understandability and computational efficiency.

Main Methods:

  • Developed Cluster-CAM, a gradient-free CNN interpretation algorithm.
  • Feature maps are split into clusters to significantly reduce the number of forward propagations required per image.
  • A strategy is employed to create cognition-based maps and cognition-scissors from clustered feature maps, which are then merged to produce the final salience heatmap.

Main Results:

  • Qualitative results demonstrate that Cluster-CAM produces heatmaps where highlighted regions align more precisely with human cognition compared to existing CAMs.
  • Quantitative evaluation confirms the superiority of Cluster-CAM in terms of both effectiveness and efficiency.
  • The algorithm significantly reduces the computational cost associated with gradient-free CAM methods.

Conclusions:

  • Cluster-CAM offers a more accurate and efficient approach to interpreting CNNs.
  • The proposed method enhances the understandability of CNN decisions in computer vision applications.
  • This technique provides a valuable tool for researchers and practitioners seeking to interpret complex deep learning models.