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

Perceptual Constancy

Perceptual Constancy

Perceptual constancy is the ability to recognize that objects remain consistent and unchanged even when their appearance varies due to changes in sensory input. There are four main types of perceptual constancy: size constancy, shape constancy, color constancy, and brightness constancy.
Size constancy is the recognition that an object remains the same size, even when its image on the retina changes. For instance, a bus is perceived to be large enough to carry people, even if it looks tiny from...

You might also read

Related Articles

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

Sort by

Same author

Chromatix: a differentiable, GPU-accelerated wave-optics library.

Nature methods·2026

Same author

Chromatix: a differentiable, GPU-accelerated wave-optics library.

bioRxiv : the preprint server for biology·2026

Same author

Pupil responses to social stimuli are associated with adaptive behaviors across the first 24 months of life.

Journal of neurodevelopmental disorders·2025

Same author

Prepronociceptin-Expressing Neurons in the Bed Nucleus of the Stria Terminalis Signal Escape Behavior.

Biological psychiatry global open science·2025

Same author

Pupil responses to social stimuli are associated with adaptive behaviors across the first 24 months of life.

bioRxiv : the preprint server for biology·2025

Same author

Motor Assisted Commutator to Harness Electronics in Tethered Experiments.

eNeuro·2025

Search research articles

Related Experiment Video

Updated: Jun 14, 2025

Quantitative Optical Microscopy: Measurement of Cellular Biophysical Features with a Standard Optical Microscope

Quantitative Optical Microscopy: Measurement of Cellular Biophysical Features with a Standard Optical Microscope

Published on: April 7, 2014

ConIQA: A deep learning method for perceptual image quality assessment with limited data.

M Hossein Eybposh^1,2, Changjia Cai^1,2, Aram Moossavi¹

¹Department of Applied Physical Sciences, The University of North Carolina at Chapel Hill, Chapel Hill, NC, 27599, USA.

Scientific Reports

|August 29, 2024

Summary

ConIQA, a novel deep learning image quality assessment method, effectively evaluates virtual and augmented reality images using consistency training. It excels in domains with limited labeled data, outperforming existing metrics for computer-generated holography.

More Related Videos

Using Computer Vision Libraries to Streamline Nuclei Quantification

Using Computer Vision Libraries to Streamline Nuclei Quantification

Published on: June 6, 2025

Area-based Image Analysis Algorithm for Quantification of Macrophage-fibroblast Cocultures

Area-based Image Analysis Algorithm for Quantification of Macrophage-fibroblast Cocultures

Published on: February 15, 2022

Related Experiment Videos

Last Updated: Jun 14, 2025

Quantitative Optical Microscopy: Measurement of Cellular Biophysical Features with a Standard Optical Microscope

Quantitative Optical Microscopy: Measurement of Cellular Biophysical Features with a Standard Optical Microscope

Published on: April 7, 2014

Using Computer Vision Libraries to Streamline Nuclei Quantification

Using Computer Vision Libraries to Streamline Nuclei Quantification

Published on: June 6, 2025

Area-based Image Analysis Algorithm for Quantification of Macrophage-fibroblast Cocultures

Area-based Image Analysis Algorithm for Quantification of Macrophage-fibroblast Cocultures

Published on: February 15, 2022

Area of Science:

Computer Vision
Image Processing
Human-Computer Interaction

Background:

Assessing image quality in VR/AR requires metrics aligned with human perception.
Deep learning image quality assessment (IQA) models struggle with domain-specific distortions and limited labeled data.
Existing FR-IQA metrics do not adequately address unique artifacts in applications like Computer-Generated Holography (CGH).

Purpose of the Study:

To develop a deep learning-based Full Reference Image Quality Assessment (FR-IQA) model, ConIQA, that overcomes limitations of existing methods in data-scarce and domain-specific scenarios.
To introduce a novel data augmentation technique and consistency training for efficient learning from both labeled and unlabeled data.
To validate ConIQA's performance on a new dataset for CGH, addressing artifacts not covered by current IQAs.

Main Methods:

Developed ConIQA, a deep learning IQA model utilizing consistency training and a novel data augmentation strategy.
Created the HQA1k dataset, containing 1000 natural images paired with CGH-rendered images, quality-rated by 13 human participants.
Evaluated ConIQA against 15 existing FR-IQA metrics on the HQA1k dataset.

Main Results:

ConIQA achieved superior performance on the HQA1k dataset, with Pearson (0.98), Spearman (0.965), and Kendall's tau (0.86) correlations.
Demonstrated significant improvements (up to 5%) in aligning with human perception compared to 15 other FR-IQA metrics.
Showcased the effectiveness of consistency training and data augmentation in low-data regimes.

Conclusions:

ConIQA offers a robust solution for image quality assessment in VR/AR, particularly in specialized domains like CGH.
The proposed method effectively learns from limited labeled data, reducing the cost and complexity of IQA model development.
ConIQA represents a significant advancement in developing perceptually relevant image quality metrics for emerging immersive technologies.