What question did this study set out to answer?

To develop a hierarchical meta-graph transformer framework that captures structural relationships among target components for robust synthetic aperture radar automatic target recognition.

August 19, 2026International Journal of Remote Sensing

HMG-Former: a hierarchical meta-graph transformer framework for robust multi-class SAR target recognition

View Full Paper

Ask AI

Bookmark

Authors

BKBibek Kumar AKAjay Kumar SGShashi Kant Gupta

Discussion

Member takes

Overview

Algorithm evaluation demonstrates high accuracy and noise robustness for HMG-Former in SAR target recognition, indicating the value of hierarchical structural modeling.

Key Points

To develop a hierarchical meta-graph transformer framework that captures structural relationships among target components for robust synthetic aperture radar automatic target recognition.
Extracted feature maps using a CNN and organized them into a hierarchical graph containing patch-level, part-level, and global target nodes.
Trained a multi-head self-attention graph transformer to learn structural dependencies and fine-grained scattering characteristics.
Evaluated the framework on the 8-class MSTAR dataset using 128 × 128 greyscale SAR images under standard conditions and speckle noise.
Achieved a 99.42% overall test accuracy and a 5-fold cross-validation accuracy of 99.39% (±0.06%).
Maintained 99.31% accuracy under speckle noise perturbation with an inference processing speed of 17.26 FPS.

Cite This Study

Kumar et al. (2026) studied this question.

synapsesocial.com/papers/6a8563d703308d306e2d735e https://doi.org/10.1080/01431161.2026.2709707

View Full Paper

Ask AI

Bookmark

August 19, 2026International Journal of Remote Sensing

HMG-Former: a hierarchical meta-graph transformer framework for robust multi-class SAR target recognition

View Full Paper

Ask AI

Bookmark

Authors

BKBibek Kumar AKAjay Kumar SGShashi Kant Gupta

Discussion

Member takes

Overview

Algorithm evaluation demonstrates high accuracy and noise robustness for HMG-Former in SAR target recognition, indicating the value of hierarchical structural modeling.

Key Points

To develop a hierarchical meta-graph transformer framework that captures structural relationships among target components for robust synthetic aperture radar automatic target recognition.
Extracted feature maps using a CNN and organized them into a hierarchical graph containing patch-level, part-level, and global target nodes.
Trained a multi-head self-attention graph transformer to learn structural dependencies and fine-grained scattering characteristics.
Evaluated the framework on the 8-class MSTAR dataset using 128 × 128 greyscale SAR images under standard conditions and speckle noise.
Achieved a 99.42% overall test accuracy and a 5-fold cross-validation accuracy of 99.39% (±0.06%).
Maintained 99.31% accuracy under speckle noise perturbation with an inference processing speed of 17.26 FPS.

Cite This Study

Kumar et al. (2026) studied this question.

synapsesocial.com/papers/6a8563d703308d306e2d735e https://doi.org/10.1080/01431161.2026.2709707

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark