Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

Reducing Line Loss

Reducing Line Loss

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...

Factorial Design

Factorial Design

Factorial Analysis is an experimental design that applies Analysis of Variance (ANOVA) statistical procedures to examine a change in a dependent variable due to more than one independent variable, also known as factors. Changes in worker productivity can be reasoned, for example, to be influenced by salary and other conditions, such as skill level. One way to test this hypothesis is by categorizing salary into three levels (low, moderate, and high) and skills sets into two levels (entry level...

Improving Translational Accuracy

Improving Translational Accuracy

Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...

Upsampling

Upsampling

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...

Maxwell-Boltzmann Distribution: Problem Solving

Maxwell-Boltzmann Distribution: Problem Solving

Individual molecules in a gas move in random directions, but a gas containing numerous molecules has a predictable distribution of molecular speeds, which is known as the Maxwell-Boltzmann distribution, f(v).
This distribution function f(v) is defined by saying that the expected number N (v1,v2) of particles with speeds between v1 and v2 is given by

You might also read

Related Articles

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

Sort by

Same author

Learning With Style: Continual Semantic Segmentation Across Tasks and Domains.

IEEE transactions on pattern analysis and machine intelligence·2024

Same author

Unsupervised Domain Adaptation of Deep Networks for ToF Depth Refinement.

IEEE transactions on pattern analysis and machine intelligence·2021

Same author

Deep Learning for Transient Image Reconstruction from ToF Data.

Sensors (Basel, Switzerland)·2021

Same author

A multi-camera dataset for depth estimation in an indoor scenario.

Data in brief·2019

Same author

Probabilistic ToF and stereo data fusion based on mixed pixels measurement models.

IEEE transactions on pattern analysis and machine intelligence·2015

Same author

Scalable coding of depth maps with R-D optimized embedding.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2013

Search research articles

Related Experiment Video

Updated: Sep 6, 2025

Gain-compensation Methodology for a Sinusoidal Scan of a Galvanometer Mirror in Proportional-Integral-Differential Control Using Pre-emphasis Techniques

Gain-compensation Methodology for a Sinusoidal Scan of a Galvanometer Mirror in Proportional-Integral-Differential Control Using Pre-emphasis Techniques

Published on: April 4, 2017

Reframing control methods for parameters optimization in adversarial image generation.

Qamar Alfalouji¹, Piergiorgio Sartor², Pietro Zanuttigh³

¹Technical University of Graz, Inffeldgasse 16b, Graz, 8010, Austria.

Neural Networks : the Official Journal of the International Neural Network Society

|June 30, 2022

Summary

This summary is machine-generated.

This study introduces Proportional-Integral (PI) and Proportional-Integral-Derivative (PID) controllers for optimizing deep learning hyper-parameters in adversarial learning. The novel strategy enhances training stability, speeds convergence, and improves image quality in models like BEGAN and CycleGAN.

Keywords:

Control methods GAN Hyper-parameters optimization

More Related Videos

Generating Strictly Controlled Stimuli for Figure Recognition Experiments

Generating Strictly Controlled Stimuli for Figure Recognition Experiments

Published on: March 18, 2019

Real-Time Proxy-Control of Re-Parameterized Peripheral Signals using a Close-Loop Interface

Real-Time Proxy-Control of Re-Parameterized Peripheral Signals using a Close-Loop Interface

Published on: May 8, 2021

Related Experiment Videos

Last Updated: Sep 6, 2025

Gain-compensation Methodology for a Sinusoidal Scan of a Galvanometer Mirror in Proportional-Integral-Differential Control Using Pre-emphasis Techniques

Gain-compensation Methodology for a Sinusoidal Scan of a Galvanometer Mirror in Proportional-Integral-Differential Control Using Pre-emphasis Techniques

Published on: April 4, 2017

Generating Strictly Controlled Stimuli for Figure Recognition Experiments

Generating Strictly Controlled Stimuli for Figure Recognition Experiments

Published on: March 18, 2019

Real-Time Proxy-Control of Re-Parameterized Peripheral Signals using a Close-Loop Interface

Real-Time Proxy-Control of Re-Parameterized Peripheral Signals using a Close-Loop Interface

Published on: May 8, 2021

Area of Science:

Deep Learning
Computer Vision
Machine Learning

Background:

Deep network training necessitates careful hyper-parameter tuning, which is critical for adversarial learning strategies in image generation.
Balancing discriminative and generative networks is essential for effective adversarial training.

Purpose of the Study:

To propose a novel hyper-parameter optimization strategy using Proportional-Integral (PI) and Proportional-Integral-Derivative (PID) controllers.
To enable efficient parameter tuning without computationally intensive trial-and-error methods.

Main Methods:

Implementation of open-loop and closed-loop schemes for tuning single or multiple hyper-parameters.
Application of the proposed PI/PID controller strategies to BEGAN and CycleGAN models.

Main Results:

Achieved more stable training and faster convergence for BEGAN and CycleGAN.
Generated sharper images with improved visual quality and better FID/FCN scores.
Image translation results demonstrated enhanced background preservation and reduced color artifacts.

Conclusions:

The proposed PI/PID controller-based hyper-parameter optimization significantly improves adversarial training efficiency and output quality.
This method offers a computationally efficient alternative to traditional trial-and-error approaches for deep learning models.