CORTEXA
← Browse
semantic_scholare-Journal of Nondestructive Testing2026-08-01

Real-Time PAUT Defect Classification with Class-Weighted Knowledge Distillation

Minsu Jeon, R. Guyon, C. Fisher, D. Mun, Jaebeom Lee

TL;DR: This work proposes a PAUT defect-classification framework that minimizes false negatives by applying class-weighted Knowledge distillation (KD) to transfer a Vision Transformer (ViT) teacher’s diagnostic capability to an extremely lightweight linear student.

Phased array ultrasonic testing (PAUT) provides high-resolution subsurface imaging through electronic beam steering and focusing and is widely used for internal defect diagnosis in structures. However, effectively interpreting PAUT data requires understanding complex spatial-temporal patterns, and deep learning has recently been introduced to address this challenge. While deep learning–based automatic interpretation has improved performance [1], most prior approaches rely on heavy CNN models that inadequately capture long-range dependencies and incur high computational cost, limiting real-time operation and on-device deployment. This work addresses these limitations by proposing a PAUT defect-classification framework that minimizes false negatives by applying class-weighted Knowledge distillation (KD) to transfer a Vision Transformer (ViT) teacher’s diagnostic capability to an extremely lightweight linear student. The proposed approach preserves high reliability while achieving real-time performance. The proposed method has three stages. First, we construct a dataset. PAUT responses are sensitive to beam angle, which can obscure or weaken defect echoes, so we pair left and right acquisitions and use them together as a two-channel input. Next, we train a ViT as the teacher model with class-weighted cross-entropy so that the model becomes more sensitive to defect classes. Finally, we transfer the teacher’s knowledge to a lightweight linear student through class-weighted KD. In this study, the lightweight linear student model consists of one convolutional layer and one linear layer. Standard KD combines a KL divergence term between teacher and student distributions with a ground-truth cross-entropy term, and the ground-truth term can reintroduce false alarms even when the teacher suppresses them. To avoid this issue and to preserve recall on defect classes, we apply the same class weights to the ground-truth term during distillation. To validate the performance of the proposed method, we compared the student model trained with the proposed scheme against ResNet and ViT models. On 937 test samples, all three models achieved an accuracy above 90%. However, while both ResNet and ViT produced two false negatives, the proposed method resulted in zero false negatives. In addition, the inference throughput was 227 images/s for ResNet and 135 images/s for ViT, whereas the proposed method achieved 2,482 images/s, reducing the inference time to less than one tenth of that of the baseline models. Per-sample inference time was about 0.0004 s, confirming real-time operation. The target computing platform for this work is an embedded inspection system under development at CEA, intended for integration with PAUT sensing hardware. While the proposed framework is ultimately envisioned to be deployed at the edge, directly on PAUT systems, the present study focuses on algorithmic feasibility rather than full system-level deployment. References [1] Guyon, R., Newson, M., Fisher, C., Miorelli, R., & Roué, D. (2025, July). Towards embedded AI models for welding defect detection in pipes. In 2025 IEEE Sensors Applications Symposium (SAS) (pp. 1-6). IEEE.

Related papers

semantic_scholare-Journal of Nondestructive Testing2026-08-01

Reducing Experimental Data Requirements in CNN-based damage detection through Transfer Learning

Finja Rentzsch holm, Tobias Schalm, Jorge Luis Jiménez Aparicio, K. Schröder

TL;DR: This study demonstrates that transfer learning enables efficient adaptation to real-world conditions, offering a cost-effective and scalable solution for data-driven SHM.

While neural networks represent a promising approach for evaluating sensor data to assess damage presence, location and severity, large amounts of data are required for training. However, the generation of experimental data is both labor-intensive and costly. Transfer learning is…

semantic_scholare-Journal of Nondestructive Testing2026-08-01

Integrated Structural Health Monitoring of Flax Fiber Reinforced Composites Using Nonlinear Resonance Acoustics, Acoustic Emission and Data-Driven Damage Identification

Othmane Achouham, C. Mechri, R. El Guerjouma, S. Allagui, Zeineb Kesentini, A. El Mahi

TL;DR: This work demonstrates that the combined use of nonlinear acoustics, acoustic emission, and machine learning constitutes a robust and highly sensitive SHM framework for composite structures.

This paper presents an integrated Structural Health Monitoring (SHM) strategy for flax fiber reinforced thermoplastic composites, combining Nonlinear Resonance Acoustic Spectroscopy (NLRAS), Acoustic Emission (AE), and data-driven damage identification based on machine learning.…

semantic_scholare-Journal of Nondestructive Testing2026-08-01

Time-Series Forecasting of Structural Temperature in Heritage Buildings Using Regression and Deep Learning Approaches

Waqas Qayyum, N. Cavalagli, E. García-Macías, F. Ubertini

Accurate prediction of the structural temperature field is crucial for the static and dynamic monitoring of engineering structures, with particular significance for heritage buildings where material preservation is paramount. The complex, time-lagged, and non-linear relationship…

semantic_scholare-Journal of Nondestructive Testing2026-08-01

Electromagnetic Assessment of Fatigue Degradation in Ferromagnetic Steel in View of Statistics and Monitoring

Christian Boller, Iman Ahadi Akhlaghi

Fatigue in metallic materials leads to progressive degradation driven by a sequence of microstructural mechanisms occurring over the life cycle. While fracture is typically the most obvious and critical damage state, it only appears at the end of life. However, when no fracture i…

semantic_scholare-Journal of Nondestructive Testing2026-08-01

Integrating Ambient Vibration Monitoring and Machine Learning for Condition Assessment of Heritage Masonry Bridges: A Venetian Case Study

Hamid Imani moghaddam, S. Russo, Raimondo Betti

Preserving the structural integrity of heritage masonry arch bridges presents unique challenges, particularly within historically dense environments like Venice where non-invasive methods are paramount. Ambient vibration monitoring (AVM) offers a well-established starting point,…