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

Distance Problem01:29

Distance Problem

When an object's velocity changes over time, the total distance traveled can be determined by summing small displacement intervals over short increments. This approach approximates the true distance through numerical summation and the use of integral calculus. An estimate of the total displacement can be obtained by measuring velocity at regular intervals and multiplying each value by the corresponding time step.If a runner accelerates over the first three seconds of a race, speed measurements...
Difference from Background: Limit of Detection01:05

Difference from Background: Limit of Detection

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...
Aggregates Classification01:29

Aggregates Classification

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...

You might also read

Related Articles

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

Sort by
Same author

ResMLP: Feedforward Networks for Image Classification With Data-Efficient Training.

IEEE transactions on pattern analysis and machine intelligence·2022
Same author

Factors of Influence for Transfer Learning Across Diverse Appearance Domains and Task Types.

IEEE transactions on pattern analysis and machine intelligence·2021
Same author

Frankenstein: Learning Deep Face Representations Using Small Data.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2017
Same author

Video2vec Embeddings Recognize Events When Examples Are Scarce.

IEEE transactions on pattern analysis and machine intelligence·2016
Same author

Interferences in Match Kernels.

IEEE transactions on pattern analysis and machine intelligence·2016
Same author

Weakly Supervised Object Localization with Multi-Fold Multiple Instance Learning.

IEEE transactions on pattern analysis and machine intelligence·2016

Related Experiment Videos

Distance-based image classification: generalizing to new classes at near-zero cost.

Thomas Mensink1, Jakob Verbeek, Florent Perronnin

  • 1University of Amsterdam.

IEEE Transactions on Pattern Analysis and Machine Intelligence
|September 21, 2013
PubMed
Summary
This summary is machine-generated.

Nearest Class Mean (NCM) classifiers offer efficient large-scale image classification, outperforming k-nearest neighbors (k-NN) and matching linear SVMs. This method efficiently incorporates new data and scales to 10,000 classes.

Related Experiment Videos

Area of Science:

  • Computer Science
  • Machine Learning
  • Artificial Intelligence

Background:

  • Large-scale image classification requires methods adaptable to continuous learning of new classes and data.
  • Existing methods like k-nearest neighbor (k-NN) and nearest class mean (NCM) have limitations in scalability and adaptability.

Purpose of the Study:

  • To develop and evaluate efficient, scalable image classification methods for continuous learning.
  • To introduce a novel metric learning approach and extensions for the NCM classifier.
  • To compare the performance of NCM against k-NN and state-of-the-art linear SVMs.

Main Methods:

  • Investigated distance-based classifiers: k-nearest neighbor (k-NN) and nearest class mean (NCM).
  • Introduced a new metric learning approach and an extension for richer class representations in NCM.
  • Conducted experiments on ImageNet 2010 (1,000 classes) and ImageNet-10K (10,000 classes) datasets.

Main Results:

  • NCM classifier demonstrated performance comparable to linear SVMs and superior to k-NN on ImageNet 2010.
  • NCM achieved competitive performance on ImageNet-10K, significantly faster than existing methods.
  • Zero-shot class priors based on ImageNet hierarchy improved performance with limited training data.

Conclusions:

  • NCM with metric learning offers a highly scalable and efficient solution for large-scale, continuous image classification.
  • The proposed NCM extensions enable richer class representations and improved generalization.
  • This approach provides a viable alternative to current state-of-the-art methods, especially for massive datasets and evolving class structures.