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

Vision

Vision

Vision is the result of light being detected and transduced into neural signals by the retina of the eye. This information is then further analyzed and interpreted by the brain. First, light enters the front of the eye and is focused by the cornea and lens onto the retina—a thin sheet of neural tissue lining the back of the eye. Because of refraction through the convex lens of the eye, images are projected onto the retina upside-down and reversed.

You might also read

Related Articles

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

Sort by

Same author

Deformation failure mechanism of deep high-stress fractured soft rock roadways and active-passive full-space collaborative control technology.

Scientific reports·2026

Same author

Tracing the stemness and malignant transition in a heritable colorectal cancer Lynch Syndrome by single-cell RNA-seq analysis.

Frontiers in immunology·2026

Same author

Association between hemoglobin glycation index and cognitive function: Evidence in the elderly.

PloS one·2026

Same author

Transcriptomic and metabolomic insights into the combined toxicity of acute glyphosate and copper exposure in water flea (Daphnia magna).

Environmental research·2026

Same author

Head-to-head comparison of fecal occult blood tests for colorectal neoplasia detection in Chinese outpatients: a multicenter diagnostic trial.

Scientific reports·2026

Same author

Non-Destructive Detection Model and Device Development for Duck Egg Freshness.

Foods (Basel, Switzerland)·2026

Search research articles

Related Experiment Video

Updated: May 10, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

Leveraging vision transformers and entropy-based attention for accurate micro-expression recognition.

Yibo Zhang^1,2, Weiguo Lin³, Yuanfa Zhang¹

¹School of Computer and Cyberspace Security, Communication University of China, Beijing, 100024, China.

Scientific Reports

|April 21, 2025

Summary

This study introduces a novel AI method for recognizing micro-expressions, improving accuracy and efficiency. The approach uses a Vision Transformer with advanced techniques to capture subtle facial cues, enhancing real-world applications.

Keywords:

Agent attention Micro-expression recognition Vision transformer

More Related Videos

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

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

Published on: April 21, 2023

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Related Experiment Videos

Last Updated: May 10, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

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

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

Published on: April 21, 2023

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Area of Science:

Computer Science
Artificial Intelligence
Machine Learning

Background:

Micro-expressions are subtle, involuntary facial movements, challenging to detect in real-time.
Existing AI systems face limitations in accuracy and efficiency due to the short duration and subtle nature of micro-expressions.
The inherent truthfulness of micro-expressions makes their accurate recognition valuable across various fields.

Purpose of the Study:

To propose a novel and effective micro-expression recognition method.
To enhance the accuracy and robustness of micro-expression recognition systems.
To address the challenges of real-time recognition and subtle facial feature detection.

Main Methods:

Development of HTNet (hierarchical transformer network with learnable absolute position embedding) for capturing subtle facial features.
Implementation of entropy-based selection agent attention to reduce model parameters and computational load.
Utilization of a diffusion model for data augmentation to increase sample size and improve generalization.

Main Results:

The proposed HTNet model demonstrates improved capacity for capturing subtle facial features.
Entropy-based attention effectively reduces model complexity while maintaining learning capability.
Diffusion-based data augmentation enhances the generalization, accuracy, and robustness of the recognition system.
Extensive experiments on multiple datasets validate the framework's effectiveness.

Conclusions:

The novel Vision Transformer-based method significantly advances micro-expression recognition.
The integrated approach of HTNet, attention mechanism, and diffusion model offers a robust solution.
The framework shows strong potential for practical applications requiring reliable micro-expression analysis.