Image-Based Time Series Representations for Pixelwise Eucalyptus Region Classification: A Comparative Study

Dias, Danielle; Dias, Ulisses; Menini, Nathalia; Lamparelli, Rubens Augusto Camargo; Maire, Guerric Le; Torres, Ricardo da S.

doi:10.1109/lgrs.2019.2946951

Cited by 25 publications

(13 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Menini et al [15], for example, investigated the use of texture descriptors based on LBP to characterize patterns of recurrence plot images constructed based of time series associated with vegetation indices. Dias et al [16], [17] approached the same problem, now exploiting different CNN-based feature extractors. All those formulations, however, were validated in the context of classification problems involving remote sensing images.…”

Section: Related Workmentioning

confidence: 99%

Non-Technical Loss Detection in Power Grid Using Information Retrieval Approaches: A Comparative Study

Esmael

Silva²,

Ji³

et al. 2021

IEEE Access

Self Cite

View full text Add to dashboard Cite

Non-technical loss (NTL) detection is a persistent challenge for Distribution System Operators. Data-driven solutions have been widely used nowadays to analyze customers' energy consumption and to identify suspicious fraud patterns for a posterior on-field inspection. However, the usage of such techniques, in particular the current deep learning methods, is not trivial and requires special attention to tackle imbalanced-class and overfitting issues. In this paper, we propose a new non-technical loss detection framework, which combines the effectiveness of convolutional neural network feature extractors with the efficiency of the Information Retrieval paradigm. In our solution, state-of-the-art pre-trained convolution neural networks (CNNs) extract deep features from electricity consumption time series represented as images. Next, these deep features are encoded into textual signatures and indexed using off-the-shelf solutions for posterior fraud searching. With this framework, the user can search for a specific fraud pattern in the utility database without having to train any classifier. The experiments performed in a real dataset provided by CPFL Energia, one of the largest electric utilities in Brazil, presented promising results both in terms of effectiveness and efficiency for the detection of fraudulent customers. In the conducted comparative study, we evaluate different time series image representations and CNN feature extraction approaches with regard to NTL detection results. Experimental results demonstrate that the combination of the Recurrence Plot image representation with the VGG16 CNN presented the best performance in terms of both effectiveness and efficiency.INDEX TERMS Content-based retrieval, deep learning, feature extraction, information retrieval, machine learning, non-technical loss detection, pattern analysis, power grids, and time series retrieval.

show abstract

Section: Related Workmentioning

confidence: 99%

Non-Technical Loss Detection in Power Grid Using Information Retrieval Approaches: A Comparative Study

Esmael

Silva²,

Ji³

et al. 2021

IEEE Access

Self Cite

View full text Add to dashboard Cite

show abstract

“…This section describes our methodology for pixelwise remote sensing image classification. Our methodology takes advantage of complementary information among image representations described in Section II, which were extracted from time series and designed for the remote sensing image classification purpose [14], [15].…”

Section: B Fusion Approachmentioning

confidence: 99%

“…With the motivation that these predictors potentially provide different views about the classified instances, we have extended previous work [14], [15] investigating a fusion method for the 28,000 predictors.…”

Section: A Evaluation Of Individual Classifiers Trained Using Deep Rmentioning

confidence: 99%

“…This section presents a comparison of performance results among different methods for fusing classifiers. In this work, we [15] 0.978 ± 0.013 0.964 ± 0.014 0.954 ± 0.012 Meta-Fusion using RP, GAF and MTF representations (accuracy) 0.982 ± 0.005 0.977 ± 0.007 0.958 ± 0.007 evaluated different approaches for fusion information available in the literature such as Bagging, Boosting, and Meta-fusion approaches: AdaBoost and Gradient Boosting are two fusion strategies that can be classified as Boosting approaches, while the Random Forest is a bagging approach by itself since several decision tree classifiers are training using the bagging approach. On the other, the Random Forest alongside with SVM algorithm can also be considered as a meta-fusion approach when they are used to have a second decision layer upon base classifiers.…”

Section: Comparison With Other Fusion Approachesmentioning

confidence: 99%

“…In part, the success of those initiatives is due to the use of effective state-of-the-art data-driven feature extractors [18]- [24]. In [14], [15], for example, ten different deep-learning feature extractors were investigated in combination with RP, GASF/GADF, and MTF representations. Remarkable results were achieved in pixelwise classification problems.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Multirepresentational Fusion of Time Series for Pixelwise Classification

Dias

Pinto

Dias

et al. 2020

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

Self Cite

View full text Add to dashboard Cite

This paper addresses the pixelwise classification problem based on temporal profiles, which are encoded in twodimensional representations based on Recurrence Plots, Gramian Angular/Difference Fields, and Markov Transition Field. We propose a multi-representational fusion scheme that exploits the complementary view provided by those time series representations, and different data-driven feature extractors and classifiers. We validate our ensemble scheme in the problem related to the classification of eucalyptus plantations in remote sensing images. Achieved results demonstrate that our proposal overcomes recently proposed baselines, and now represents the new state-of-the-art classification solution for the target dataset.

show abstract

Behavioral Biometrics Authentication in Critical Infrastructure Using Siamese Neural Networks

Budžys

Kurasova

Medvedev

2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Image-Based Time Series Representations for Pixelwise Eucalyptus Region Classification: A Comparative Study

Cited by 25 publications

References 18 publications

Non-Technical Loss Detection in Power Grid Using Information Retrieval Approaches: A Comparative Study

Non-Technical Loss Detection in Power Grid Using Information Retrieval Approaches: A Comparative Study

A Multirepresentational Fusion of Time Series for Pixelwise Classification

Behavioral Biometrics Authentication in Critical Infrastructure Using Siamese Neural Networks

Contact Info

Product

Resources

About