Table Predictive Preformance Results - 3D Joint Spatiotemporal Vision Transformer-based Approach for Defect Segmentation using Active Infrared Thermography

Name: Table Predictive Preformance Results - 3D Joint Spatiotemporal Vision Transformer-based Approach for Defect Segmentation using Active Infrared Thermography
Creator: Adrián De-Miguel-Palacio
Published: 2026-09-30T08:44:45.477Z
Keywords: Material Defect Analysis, Active Thermography

De-Miguel-Palacio, Adrián

doi:10.17632/jg258c7cp7.1

Table Predictive Preformance Results - 3D Joint Spatiotemporal Vision Transformer-based Approach for Defect Segmentation using Active Infrared Thermography

Published: 30 September 2026| Version 1 | DOI: 10.17632/jg258c7cp7.1

Contributor:

Adrián De-Miguel-Palacio

Description

CSV with split-wise evaluation results (five folds) used to compute the predictive metrics (IoU, F1, POD, and FPR) for DS-AIRT 3D Swin UNETR and the baseline methods. These values were used to obtain the mean and standard deviation, as well as to perform the one-tailed paired t-tests, statistical power analysis, and Cohen’s effect-size calculations reported in the table.

Table Predictive Preformance Results - 3D Joint Spatiotemporal Vision Transformer-based Approach for Defect Segmentation using Active Infrared Thermography

Description

Files

Categories

Licence