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

Search research articles

Related Experiment Video

Updated: Jan 15, 2026

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

A multiscale transformer with spatial attention for hyperspectral image classification.

Irfan Ahmad¹, Ghulam Farooque², Fazal Hadi³

¹School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, 210094, Jiangsu, China.

Scientific Reports

|January 13, 2026

Summary

Related Concept Videos

You might also read

Related Articles

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

Sort by

Same author

An effective approach to improving photovoltaic defect detection using the new DCD-YOLOv8s model.

Scientific reports·2025

Same author

Patterns and trends in prevalent central nervous system stimulant use in US veterans with traumatic brain injury.

The American journal of medicine·2025

Same author

Barriers and facilitators of antiseizure medication adherence: a qualitative study among persons with epilepsy in Pakistan.

BMC neurology·2025

Same author

Medication therapy management in Pakistan: a cross-sectional evaluation of pharmacists' knowledge, attitudes, practices, and barriers.

Journal of pharmaceutical health care and sciences·2025

Same author

Periocular clinical differences between primary and postparalytic hemifacial spasm: A retrospective comparative study.

Indian journal of ophthalmology·2025

Same author

Integrating socio-behavioral, ethics, community, and translational science considerations in HIV cure research.

Current opinion in HIV and AIDS·2025

Same journal

Correction: A method for supervoxel-wise association studies of age and other non-imaging variables from coronary computed tomography angiograms.

Scientific reports·2026

Same journal

Poly(bromophenol blue)/CoSn(OH)<sub>6</sub> cubic particles modified pencil graphite electrode for electrochemical determination of diphenhydramine.

Scientific reports·2026

Same journal

Dietary Chlorella, Spirulina, and acidifier modulate jejunal cytokine-related gene expression in broiler chickens.

Scientific reports·2026

Same journal

Perceived physical activity barriers in university students: associations with fatigue and eating behaviours.

Scientific reports·2026

Same journal

Refuge limitation structures habitat use in agricultural landscapes: evidence from Sunda pangolins.

Scientific reports·2026

Same journal

Lightweight stateless transaction verification with outsourced witness updates for UTXO blockchains.

Scientific reports·2026

See all related articles

This summary is machine-generated.

MTSA-Net enhances hyperspectral image (HSI) classification by integrating multiscale transformers and spatial attention, outperforming existing methods, especially with limited training data.

Area of Science:

Remote Sensing
Computer Vision
Machine Learning

Background:

Hyperspectral images (HSIs) offer rich spectral-spatial information vital for accurate classification.
Convolutional Neural Networks (CNNs) show promise but face limitations in capturing long-range dependencies and generalization due to fixed scales.
Network depth in CNNs can paradoxically lead to performance degradation in HSI classification.

Purpose of the Study:

To introduce MTSA-Net, a novel framework for hyperspectral image classification.
To address the limitations of CNNs in capturing long-range dependencies and improving generalization.
To develop a robust and flexible HSI classification approach using multiscale transformers and spatial attention.

Main Methods:

Utilizes 3-D and 2-D convolution layers followed by spatial attention to focus on critical spatial features.

Keywords:

Convolutional neural networks (CNNs)Hyperspectral image classification Multiscale transformers Spatial attention Spectral-spatial features

Related Experiment Videos

Last Updated: Jan 15, 2026

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

Employs multiscale transformer encoders to capture both local and global representations and model long-range dependencies.

Integrates a feature fusion module to combine features from varying scales for comprehensive representation.

Main Results:

MTSA-Net achieves superior performance compared to state-of-the-art methods on five benchmark HSI datasets.
Achieved high overall accuracies: 98.84% (Indian Pines), 98.77% (Pavia University), 99.80% (Salinas Valley), 97.84% (Houston-13), and 95.87% (Houston-18).
Demonstrates significant effectiveness, particularly in scenarios with limited training samples.

Conclusions:

MTSA-Net offers a robust, flexible, and high-performing framework for hyperspectral image classification.
The integration of multiscale transformers and spatial attention effectively captures spectral-spatial features and long-range dependencies.
The proposed method shows strong potential for advancing HSI classification, especially under data-scarce conditions.