What type of study is this?

September 14, 2025

Leveraging modified ex situ tomography data for segmentation of in situ synchrotron X-ray computed tomography.

View Full Paper

Ask AI

Bookmark

Authors

Methodology demonstrates automated segmentation in complex imaging with 94.7% IoU, indicating robust performance across materials.

The proposed model achieves 94.7% Intersection over Union (IoU) for segmentation, matching human reliability at 94.6% IoU.
By employing a modified SegFormer architecture, the model drastically reduces processing time by 100 times for 3D datasets.
Despite training only on static specimens, the approach shows robust performance through significant morphological changes during analyses.
Applicable to varied materials systems, this method enhances the assessment of time-resolved tomographic data across scientific fields.

Manchester et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark