A Hybrid Deep Learning Framework for Unsupervised Anomaly Detection in Multivariate Spatio-Temporal Data

Karadayı, Yıldız; Aydın, Mehmet N.; Öğrenci, Arif Selçuk

doi:10.3390/app10155191

Cited by 24 publications

(7 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When an input image passed through a standard neural network, many of the temporal [29] (time-related: pictures that were taken at different time intervals) and spatial [30] (spacerelated: properties related to a single image such as coordinates, gradients, resolution and so on) features are lost. Convolutional Neural Network -ConvNet -CNN [31] model is used to overcome this problem.…”

Section: A Training a Deep Neural Network Through Convnetsmentioning

confidence: 99%

Anomaly Detection on Medical Images using Autoencoder and Convolutional Neural Network

Siddalingappa¹,

Sekar²

2021

IJACSA

View full text Add to dashboard Cite

Detection of anomalies from the medical image dataset improves prognosis by discovering new facts hidden in the data. The present study aims to discuss anomaly detection using autoencoders and convolutional neural networks. The autoencoder identifies the imbalance between normal and abnormal samples. They create learning models flexible and accurate on training data. The problem is addressed in four stages: 1) training: an autoencoder is initialized with the hyperparameters and trained on the lung cancer CT scan images, 2) test: the autoencoder reconstructs the input from the latent space representation with a slight variation from the original data, indicated by a reconstruction error as Mean Squared Error (MSE), 3) evaluate: the MSE value of the training and test dataset are compared. The MSE values of anomalous data are higher than a base threshold, detecting those as anomalies, 4) validate: the efficiency metrics such as accuracy and MSE scores are used at both training and validation phases. The dataset was further classified as benign and malignant. The accuracy reported for outlier detection and the classification task is 98% and 97.2%. Thus, the proposed autoencoder-based anomaly detection could positively isolate anomalies from the CT scan images of lung cancer.

show abstract

Section: A Training a Deep Neural Network Through Convnetsmentioning

confidence: 99%

Anomaly Detection on Medical Images using Autoencoder and Convolutional Neural Network

Siddalingappa¹,

Sekar²

2021

IJACSA

View full text Add to dashboard Cite

show abstract

“…21,22 The recent advancement of deep learning-based anomaly detection methods have gained much popularity with their promising performance. 23,24 For instance, implementing the anomaly detection using the Auto-Encoder (AE) by inspecting its reconstruction errors. 25 Additionally, Recurrent Neural Networks (RNNs) architectures and have resulted in outstanding performance for a variety of problems including time series prediction and sequence-to-sequence learning.…”

Section: Related Workmentioning

confidence: 99%

Comparison of deep learning-based methods in multimodal anomaly detection: A case study in human–robot collaboration

Yang

Yan

2021

Science Progress

View full text Add to dashboard Cite

Human-Robot Collaboration (HRC) has been widely used in daily life and industry for maximizing the advantages of humans and robots, respectively. However, the internal modeling errors or external perturbations still affect robotic systems such as human collisions and environmental changes. Multimodal anomaly detection plays an increasingly important role in HRC applications, which detects unexpected anomalies from multimodal signals. Due to the complex temporal dependence and stochasticity, it is still difficult to choose a common model applicable to all collaborative tasks, and lack of comparative analysis of existing methods and verification of specific application cases. In this paper, six representative deep learning-based methods are evaluated and the comparing metrics including detection accuracy, multi-modality combinations, and anomaly time bias. For a fair comparison, each detector models multimodal signals from non-anomalous samples and then determines an anomaly using a predefined threshold. We evaluate the detectors with force, torque, velocity, tactile, and kinematic sensing during a human-robot kitting experiment that consists of six individual skills, results indicate that the LSTM-DAGMM based detector outperformed the others, which yielding higher accuracy and efficiency. The metrics are measured with the RUC and ROC by changing the settings of multi-modality combinations and various anomaly biases, which aim to obtain the best performance of multimodal anomaly detection.

show abstract

“…However, the performance of such methods significantly degrades for high-dimensional data or in the presence of noise. To overcome such limitations deep learning solutions [49][50][51][52][53], have shown to be very effective in solving unsupervised anomaly detection problems.…”

Section: Related Workmentioning

confidence: 99%

Online Learning of Oil Leak Anomalies in Wind Turbines with Block-Based Binary Reservoir

et al. 2021

View full text Add to dashboard Cite

The focus of this work is to design a deeply quantized anomaly detector of oil leaks that may happen at the junction between the wind turbine high-speed shaft and the external bracket of the power generator. We propose a block-based binary shallow echo state network (BBS-ESN) architecture belonging to the reservoir computing (RC) category and, as we believe, it also extends the extreme learning machines (ELM) domain. Furthermore, BBS-ESN performs binary block-based online training using fixed and minimal computational complexity to achieve low power consumption and deployability on an off-the-shelf micro-controller (MCU). This has been achieved through binarization of the images and 1-bit quantization of the network weights and activations. 3D rendering has been used to generate a novel publicly available dataset of photo-realistic images similar to those potentially acquired by image sensors on the field while monitoring the junction, without and with oil leaks. Extensive experimentation has been conducted using a STM32H743ZI2 MCU running at 480 MHz and the results achieved show an accurate identification of anomalies, with a reduced computational cost per image and memory occupancy. Based on the obtained results, we conclude that BBS-ESN is feasible on off-the-shelf 32 bit MCUs. Moreover, the solution is also scalable in the number of image cameras to be deployed and to achieve accurate and fast oil leak detections from different viewpoints.

show abstract

A Hybrid Deep Learning Framework for Unsupervised Anomaly Detection in Multivariate Spatio-Temporal Data

Cited by 24 publications

References 42 publications

Anomaly Detection on Medical Images using Autoencoder and Convolutional Neural Network

Anomaly Detection on Medical Images using Autoencoder and Convolutional Neural Network

Comparison of deep learning-based methods in multimodal anomaly detection: A case study in human–robot collaboration

Online Learning of Oil Leak Anomalies in Wind Turbines with Block-Based Binary Reservoir

Contact Info

Product

Resources

About