Published on Fri Oct 04 2019

AKMD : An Adaptive Framework for Online Sensing and Anomaly Quantification

Hao Yan, Kamran Paynabar, Jianjun Shi

Adaptive sensing through a systematic exploration is vital for online inspection and anomaly quantification. Most of the existing sequential sampling methodologies focus on reducing the overall fitting error. The proposed methodology balances the sampling efforts between the space-filling sampling and focused sampling near the anomalous region.

0
0
0
Abstract

In point-based sensing systems such as coordinate measuring machines (CMM) and laser ultrasonics where complete sensing is impractical due to the high sensing time and cost, adaptive sensing through a systematic exploration is vital for online inspection and anomaly quantification. Most of the existing sequential sampling methodologies focus on reducing the overall fitting error for the entire sampling space. However, in many anomaly quantification applications, the main goal is to estimate sparse anomalous regions in the pixel-level accurately. In this paper, we develop a novel framework named Adaptive Kernelized Maximum-Minimum Distance AKMD to speed up the inspection and anomaly detection process through an intelligent sequential sampling scheme integrated with fast estimation and detection. The proposed method balances the sampling efforts between the space-filling sampling (exploration) and focused sampling near the anomalous region (exploitation). The proposed methodology is validated by conducting simulations and a case study of anomaly detection in composite sheets using a guided wave test.

Fri Jul 19 2013
Computer Vision
Automated Defect Localization via Low Rank Plus Outlier Modeling of Propagating Wavefield Data
This work proposes an agnostic inference strategy for material diagnostics. The proposed approach couples spatiotemporal windowing with low rank plus outlier modeling, to identify a priori unknown defect.
0
0
0
Sun Apr 15 2018
Computer Vision
Detecting Concrete Abnormality Using Time-series Thermal Imaging and Supervised Learning
Nondestructive detecting defects (NDD) in concrete structures have been explored for decades. The major limitations are the high noises to signal created from the environmental factors. Future studies will focus on implementing more sophisticated structured models to handle more realistic cases.
0
0
0
Fri Aug 27 2021
Computer Vision
Anomaly Detection of Defect using Energy of Point Pattern Features within Random Finite Set Framework
Most recent works use global features for feature extraction. Global features are not robust against lighting and viewpoint changes. We are the first to propose using transfer learning of local/point pattern features. We model these features as a random finite set.
1
0
0
Mon Dec 21 2020
Machine Learning
A Bayesian methodology for localising acoustic emission sources in complex structures
Acoustic emissions to localise damage sources has emerged as a popular approach. Despite recent advances, the task of locating damage within composite material still poses a significant challenge. A Bayesian source localisation strategy that is robust to these complexities is presented.
0
0
0
Sat Nov 10 2018
Machine Learning
Anomaly Detection via Graphical Lasso
Anomalies and outliers are common in real-world data, and they can arise from many sources, such as sensor faults. Standard Graphical Lasso-based techniques can identify the conditional dependency structure of a collection of random variables based on their sample covariance matrix. In particular, several outliers in a sample covarianance matrix can destroy the sparse inverse of
0
0
0
Fri May 07 2021
Machine Learning
Energy-Based Anomaly Detection and Localization
This brief sketches initial progress towards a unified energy-based solution for the semi-supervised visual anomaly detection and localization problem. In this setup, we have access to only anomaly-free training data and want to identify anomalies of an arbitrary nature on test data.
2
0
0