k-Shape

Paparrizos, John; Gravano, Luis

doi:10.1145/2949741.2949758

Cited by 137 publications

(28 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Regarding the GPR data processing and interpretation, follow-up works include the use of GPR phase analysis techniques [30] as an additional tool for the identification of cavities in corroded concrete. We also plan to explore the use of time-series clustering algorithms [31] to analyze profile traces in order to automate the identification and categorization of key features (such as cavities, rebars, and moist areas).…”

Section: Discussionmentioning

confidence: 99%

Assessing Rebar Corrosion through the Combination of Nondestructive GPR and IRT Methodologies

Solla

Lagüela

Fernández³

et al. 2019

Remote Sensing

View full text Add to dashboard Cite

Corrosion is one of the pathologies that most affects the resistance of reinforced concrete. There are numerous ancient structures still in use affected by corrosion that need proper evaluation and remedial treatment for their maintenance. In this sense, there has been an increasing tendency to use nondestructive testing techniques that do not alter the reinforcement elements of such vulnerable structures. This work presents a combined methodology by using ground penetrating radar (GPR) and infrared thermography (IRT) techniques for the detection and evaluation of corrosion. The methodology was applied to the case study of an old construction that belongs to the abandoned military battery of Cabo Udra (Galicia, Spain). The combination of these complementary techniques allowed for the identification of areas with different dielectric and thermal conductivity, as well as different reflection patterns and intensity of the GPR waves. Thus, from the analysis of the GPR signals and IRT images acquired, it was possible to interpret corroded areas and moisture, along with inner damages such as cracking and debonding. These pathologies have a direct effect on the durability and sustentation of a structure, while the knowledge of their existence might be useful for engineers engaged in the design of maintenance works.

show abstract

Section: Discussionmentioning

confidence: 99%

Assessing Rebar Corrosion through the Combination of Nondestructive GPR and IRT Methodologies

Solla

Lagüela

Fernández³

et al. 2019

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Time series clustering methods group training time series into a number of (typically prespecified) clusters. A popular choice in time series analysis is k-means, being the only clustering algorithm scaling linearly with data set size [6]. Here, DTW-based elastic measures might again be used to find the barycenter best representing the centroids of k-means clusters [19].…”

Section: Methods Based On One-dimensional Representationsmentioning

confidence: 99%

“…Multidimensional representations: These representations emerge when the sensor data are projected to a dual space by extraction of features. When aiming for a generic anomaly detection model, the major challenge is given by the choice of a generic but expressive set of features [6]. Among popular choices are statistical measures and wavelet-based features [7] or filter bank features (e.g., Mel-frequency cepstral coefficient (MFCC) features) [8].…”

mentioning

confidence: 99%

Collecting Labels for Rare Anomalies via Direct Human Feedback—An Industrial Application Study

2019

View full text Add to dashboard Cite

Many systems rely on the expertise from human operators, who have acquired their knowledge through practical experience over the course of many years. For the detection of anomalies in industrial settings, sensor units have been introduced to predict and classify such anomalous events, but these critically rely on annotated data for training. Lengthy data collection campaigns are needed, which tend to be combined with domain expert annotations of the data afterwards, resulting in costly and slow process. This work presents an alternative by studying live annotation of rare anomalous events in sensor streams in a real-world manufacturing setting by experienced human operators that can also observe the machinery itself. A prototype for visualization and in situ annotation of sensor signals is developed with embedded unsupervised anomaly detection algorithms to propose signals for annotation and which allows the operators to give feedback on the detection and classify anomalous events. This prototype allowed assembling a corpus of several weeks of sensor data measured in a real manufacturing surrounding and was annotated by domain experts as an evaluation basis for this study. The evaluation of live annotations reveals high user motivation after getting accustomed to the labeling prototype. After this initial period, clear anomalies with characteristic signal patterns are detected reliably in visualized envelope signals. More subtle signal deviations were less likely to be confirmed an anomaly due to either an insufficient visibility in envelope signals or the absence of characteristic signal patterns.

show abstract

“…This method uses a normalized cross-correlation function to derive a shape-based distance metric 27 . The tslearn implementation is used in Mesmerize 28 .…”

Section: K-shape Clusteringmentioning

confidence: 99%

“…Mesmerize also provides an implementation for k-Shape clustering 27,28 to extract a finite set of discrete archetypical peaks from calcium traces (Fig 2i). These archetypical peaks allow traces to be reduced to sequences of discrete letters that can be modeled with techniques such as Markov Chains (Fig 2j-m).…”

mentioning

confidence: 99%

Mesmerize: a dynamically adaptable user-friendly analysis platform for 2D & 3D calcium imaging data

Kolar

Dondorp

Zwiggelaar

et al. 2019

Preprint

View full text Add to dashboard Cite

We present an efficient and expandable calcium imaging analysis platform that encapsulates the entire analysis process from raw data to interactive e-figures. It provides a graphical interface to the latest analysis methods for pre-processing, and signal extraction. We demonstrate how Mesmerize can be applied to a broad range of scientific questions by using datasets ranging from the mouse visual cortex, neurons, epidermis and TLCs of the protochordate C. intestinalis, and C. elegans.Large-scale calcium imaging of neuronal activity in populated brain regions, or entire animals, has become an indispensable technique in neuroscience research. The analysis of calcium imaging datasets presents significant challenges with respect to signal extraction and data organization.Numerous commendable packages, such as the Caiman library 1 , Suite2p 2 and SIMA 3 have been created to address pre-processing and signal extraction of imaging data. However, users often incorporate these tools in custom written scripts without a system to link the analysis procedures, raw and analyzed data.This non-ideal organization can greatly impede the reproducibility of the work even when the raw data are available 4,5 . Projects such as the Jupyter notebook 6 can help address some of these issues; however these platforms are not easily accessible to non-programmers. Moreover, these solutions do not have a system in place for FAIR, open & functionally linked data 4,7,8 . In order to address these organizational

show abstract

k-Shape

Cited by 137 publications

References 59 publications

Assessing Rebar Corrosion through the Combination of Nondestructive GPR and IRT Methodologies

Assessing Rebar Corrosion through the Combination of Nondestructive GPR and IRT Methodologies

Collecting Labels for Rare Anomalies via Direct Human Feedback—An Industrial Application Study

Mesmerize: a dynamically adaptable user-friendly analysis platform for 2D & 3D calcium imaging data

Contact Info

Product

Resources

About