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

Long-patch Base Excision Repair01:02

Long-patch Base Excision Repair

8.0K
Since the discovery of the two BER pathways, there has been a debate about how a cell chooses one pathway over the other and the factors determining this selection. Numerous in vitro experiments have pointed out multiple determinants for the sub-pathway selection. These are:
8.0K
Bacterial Transformation01:33

Bacterial Transformation

59.7K
In 1928, bacteriologist Frederick Griffith worked on a vaccine for pneumonia, which is caused by Streptococcus pneumoniae bacteria. Griffith studied two pneumonia strains in mice: one pathogenic and one non-pathogenic. Only the pathogenic strain killed host mice.
Griffith made an unexpected discovery when he killed the pathogenic strain and mixed its remains with the live, non-pathogenic strain. Not only did the mixture kill host mice, but it also contained living pathogenic bacteria that...
59.7K
Transformation01:26

Transformation

883
Microbial communities are dynamic environments where cell lysis releases free DNA into the surroundings. Other cells can take up this extracellular DNA through a process known as transformation.When a cell incorporates this foreign DNA into its genome, resulting in genetic modification, the process is known as transformation. Cells capable of this process are termed competent. Competence can be natural, as observed in certain bacteria and archaea, or artificially induced in the...
883
Transformers01:26

Transformers

1.8K
A device that transforms voltages from one value to another using induction is called a transformer. A transformer consists of two separate coils, or windings, wrapped around the same soft iron core. However, they are electrically insulated from each other.
The iron core has a substantial relative permeability. Therefore, the magnetic field lines generated due to the current in one winding are almost entirely confined within the core, such that the same magnetic flux permeates each turn of both...
1.8K
Avoidance Learning and Learned Helplessness01:14

Avoidance Learning and Learned Helplessness

2.6K
Avoidance learning and learned helplessness are critical concepts in understanding behavioral responses to negative stimuli.
Avoidance learning occurs when an organism learns that a specific behavior can prevent an unpleasant outcome. For example, a student who receives a bad grade may start studying harder to avoid future poor grades. This behavior persists even when the negative outcome is no longer present. Avoidance learning is powerful because it maintains behavior in the absence of the...
2.6K
Behavior of Concrete Under Compressive Load01:23

Behavior of Concrete Under Compressive Load

618
Concrete exhibits specific behaviors under different compressive loads. Understanding this is crucial for understanding its structural integrity. When concrete undergoes uniaxial compression, it tends to develop cracks that run parallel to the direction of the force. These parallel cracks stem from localized tensile stresses that occur perpendicular to the compression direction. Additionally, angled cracks may appear due to the formation of shear planes.
As the concrete specimen fractures under...
618

You might also read

Related Articles

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

Sort by
Same author

Enhancing YOLOv5 for Autonomous Driving: Efficient Attention-Based Object Detection on Edge Devices.

Journal of imaging·2025
Same author

Road-Type Classification with Deep AutoEncoder.

Computational intelligence and neuroscience·2023
Same author

Automatic Dynamic Range Adjustment for Pedestrian Detection in Thermal (Infrared) Surveillance Videos.

Sensors (Basel, Switzerland)·2022
Same author

Spatio-Temporal Scale Coded Bag-of-Words.

Sensors (Basel, Switzerland)·2020

Related Experiment Video

Updated: Jan 28, 2026

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.3K

Patched-Based Swin Transformer Hyperprior for Learned Image Compression.

Sibusiso B Buthelezi1, Jules R Tapamo1

  • 1Discipline of Electrical, Electronic and Computer Engineering, University of KwaZulu-Natal, Durban 4041, South Africa.

Journal of Imaging
|January 27, 2026
PubMed
Summary

This study introduces a hybrid learned image compression method using a CNN-based variational autoencoder (VAE) and a Swin Transformer. The novel approach enhances compression efficiency and visual quality by effectively modeling global dependencies while maintaining computational feasibility.

Keywords:
differentiable quantization modulelatent variable modellingrate-distortion optimizationswin transformer hyperpriorvariational inference

More Related Videos

Constructing and Visualizing Models using Mime-based Machine-learning Framework
06:19

Constructing and Visualizing Models using Mime-based Machine-learning Framework

Published on: July 22, 2025

2.4K
Slice Patch Clamp Technique for Analyzing Learning-Induced Plasticity
11:56

Slice Patch Clamp Technique for Analyzing Learning-Induced Plasticity

Published on: November 11, 2017

16.3K

Related Experiment Videos

Last Updated: Jan 28, 2026

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.3K
Constructing and Visualizing Models using Mime-based Machine-learning Framework
06:19

Constructing and Visualizing Models using Mime-based Machine-learning Framework

Published on: July 22, 2025

2.4K
Slice Patch Clamp Technique for Analyzing Learning-Induced Plasticity
11:56

Slice Patch Clamp Technique for Analyzing Learning-Induced Plasticity

Published on: November 11, 2017

16.3K

Area of Science:

  • Computer Vision
  • Machine Learning
  • Image Processing

Background:

  • Traditional learned image compression methods using Convolutional Neural Networks (CNNs) struggle with capturing long-range dependencies in latent representations due to localized receptive fields, limiting compression efficiency.
  • Fully transformer-based models offer global dependency modeling but incur high computational costs, making them unsuitable for high-resolution image compression.

Purpose of the Study:

  • To develop a hybrid end-to-end learned image compression framework that overcomes the limitations of existing entropy models.
  • To effectively model both local and global contextual information within computational constraints for improved image compression.

Main Methods:

  • A hybrid framework combining a CNN-based variational autoencoder (VAE) with a patch-based hierarchical Swin Transformer hyperprior.
  • Utilizing shifted window self-attention in the Swin Transformer to capture local and global dependencies efficiently.
  • Integrating the entropy model with a differentiable quantization module for end-to-end joint optimization of the rate-distortion objective.

Main Results:

  • The proposed hybrid architecture achieves superior rate-distortion performance on standard datasets (Kodak, JPEG AI, CLIC).
  • Demonstrates enhanced compression performance with higher visual quality at lower bitrates compared to CNN-based entropy priors.
  • The model learns a more accurate probability distribution of the latent representation, improving bitrate estimation and enabling more compact latent representations.

Conclusions:

  • Integrating efficient transformer architectures, like the Swin Transformer, into learned image compression is effective.
  • The hybrid approach offers a viable solution for advanced entropy modeling beyond conventional CNN-based designs.
  • This work advances the field of learned image compression by balancing performance and computational efficiency.