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

Transformers with Off-Nominal Turns Ratios01:25

Transformers with Off-Nominal Turns Ratios

204
In scenarios involving parallel transformers with disparate ratings, developing per-unit models requires accommodating off-nominal turns ratios. This situation arises when the selected base voltages are not proportional to the transformer’s voltage ratings. Consider a transformer where the rated voltages are related by the term a. If the chosen voltage bases satisfy a relationship involving term b, term c is defined as the ratio of these bases. This ratio is then substituted into the...
204
Reducing Line Loss01:18

Reducing Line Loss

191
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...
191
Energy Losses in Transformers01:21

Energy Losses in Transformers

953
In an ideal transformer, it is assumed that there are no energy losses, and, hence, all the power at the primary winding is transferred to the secondary winding. However, in reality,  the transformers always have some energy losses, and, hence, the output power obtained at the secondary winding is less than the input power at the primary winding due to energy losses.
There are four main reasons for energy losses in transformers.
The first cause can be  the high resistance of the...
953
Three-Winding Transformers01:19

Three-Winding Transformers

303
Three identical single-phase transformers can be configured to form a three-phase transformer connection, which involves high-voltage and low-voltage windings. The high-voltage windings are denoted by capital letters A-B-C, while the low-voltage windings are labeled with lowercase letters a-b-c, representing their respective phases. This notation helps distinguish between the high and low voltage sides of the transformer.
In the per-unit equivalent circuit of a grounded Y-Y three-phase...
303
Types Of Transformers01:16

Types Of Transformers

1.0K
Transformers can provide desired voltages to a circuit by modifying the number of turns in the secondary windings.
If the ratio of the number of turns in the secondary winding to that of the primary winding is greater than one, then the transformer is said to be a step-up transformer. In a step-up transformer, the voltage at the secondary winding is greater than the voltage applied at the primary winding.
However, if this ratio is less than one, the transformer is said to be a step-down...
1.0K
Equivalent Circuits for Practical Transformers01:28

Equivalent Circuits for Practical Transformers

783
The practical equivalent circuits of single-phase two-winding transformers exhibit significant deviations from their idealized versions due to the inherent properties of winding resistance and finite core permeability. These properties result in real and reactive power losses, affecting the transformer's performance. Understanding these deviations is crucial for designing more efficient transformers.
In a practical transformer, each winding exhibits resistance and leakage reactance. The...
783

You might also read

Related Articles

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

Sort by
Same author

Impact of labelling inaccuracy and image noise on tooth segmentation in panoramic radiographs using federated, centralized, and local learning.

Dento maxillo facial radiology·2026
Same author

Continual low-rank scaled dot-product attention.

Neural networks : the official journal of the International Neural Network Society·2025
Same author

A multimodal stress detection dataset with facial expressions and physiological signals.

Scientific data·2025
Same author

Manifold Gaussian Variational Bayes on the Precision Matrix.

Neural computation·2024
Same author

A synthetic data set to benchmark anti-money laundering methods.

Scientific data·2023
Same author

Automatic social distance estimation for photographic studies: Performance evaluation, test benchmark, and algorithm.

Machine learning with applications·2022

Related Experiment Video

Updated: Sep 5, 2025

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

1.9K

Single-layer vision transformers for more accurate early exits with less overhead.

Arian Bakhtiarnia1, Qi Zhang1, Alexandros Iosifidis1

  • 1DIGIT, Department of Electrical and Computer Engineering, Aarhus University, Denmark.

Neural Networks : the Official Journal of the International Neural Network Society
|July 11, 2022
PubMed
Summary

This study presents a new early exiting method for deep learning models, enhancing accuracy and efficiency in resource-limited edge computing and IoT systems. The approach improves dynamic inference for time-critical applications.

Keywords:
Dynamic inferenceEarly exitingMulti-exit architectureMultimodal deep learningVision transformer

More Related Videos

Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping
07:11

Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping

Published on: December 8, 2023

1.9K
Assessing Binocular Central Visual Field and Binocular Eye Movements in a Dichoptic Viewing Condition
07:45

Assessing Binocular Central Visual Field and Binocular Eye Movements in a Dichoptic Viewing Condition

Published on: July 21, 2020

4.5K

Related Experiment Videos

Last Updated: Sep 5, 2025

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

1.9K
Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping
07:11

Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping

Published on: December 8, 2023

1.9K
Assessing Binocular Central Visual Field and Binocular Eye Movements in a Dichoptic Viewing Condition
07:45

Assessing Binocular Central Visual Field and Binocular Eye Movements in a Dichoptic Viewing Condition

Published on: July 21, 2020

4.5K

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Deploying deep learning models in resource-constrained environments like edge computing and IoT networks presents significant challenges for time-critical applications.
  • Dynamic inference methods, particularly early exiting, are crucial for managing computational load and latency in such scenarios.

Purpose of the Study:

  • To introduce a novel early exiting architecture based on the vision transformer.
  • To develop a fine-tuning strategy that enhances the accuracy of early exit branches with reduced overhead.
  • To demonstrate the method's versatility across different data modalities and task types.

Main Methods:

  • A new early exiting architecture integrated with the vision transformer model.
  • A specialized fine-tuning strategy to boost the performance of early exit branches.
  • Extensive experimentation on image classification, audio classification, and audiovisual crowd counting tasks.

Main Results:

  • The proposed method significantly improves the accuracy of early exit branches compared to conventional approaches.
  • The architecture introduces less computational overhead, making it suitable for edge devices.
  • The approach demonstrates effectiveness in both single-modal (image, audio) and multi-modal (audiovisual) settings.
  • A novel method for integrating audio-visual modalities in early exits was introduced, enabling finer-grained dynamic inference.

Conclusions:

  • The novel early exiting strategy offers a more accurate and efficient solution for deploying deep learning models in time-critical, resource-limited applications.
  • The method's adaptability to various tasks (classification, regression) and data types (single- and multi-modal) highlights its broad applicability.
  • The integrated audiovisual early exit mechanism advances dynamic inference capabilities for complex data analysis.