scispace - formally typeset
Search or ask a question
Proceedings ArticleDOI

Knowledge-Graph Based Multi-Target Deep-Learning Models for Train Anomaly Detection

TL;DR: By establishing a knowledge graph of the normally operating systems, a progressive approach to use a multi-target network to segment each component of the considered system sequentially by decoupling the segmentation and the classification task is proposed.
Abstract: The state-of-art image segmentation algorithms can be applied to accurately localize objects by using deep convolutional neural networks (CNN). In this paper, we consider the anomaly detection problem encountered in a train wheel system. We propose a progressive approach to use a multi-target network to segment each component of the considered system sequentially by decoupling the segmentation and the classification task. Moreover, we use the knowledge graph approach to establish a semantic consistency matrix by quantifying the spatial relationship between various components. We show that by establishing a knowledge graph of the normally operating systems, we are able to identify a faulty component effectively.
Citations
More filters
Book ChapterDOI
16 Dec 2019
TL;DR: This article will provide an overview of the fault KG and summarize the key techniques for building a KG to guide the construction of the KG in the fault domain, and introduces some of the open source tools that can be used to build a KGs process, enabling researchers and practitioners to quickly get started in this field.
Abstract: Knowledge Graph (KG) is a graph-based data structure that can display the relationship between a large number of semi-structured and unstructured data, and can efficiently and intelligently search for information that users need. KG has been widely used for many fields including finance, medical care, biological, education, journalism, smart search and other industries. With the increase in the application of Knowledge Graphs (KGs) in the field of failure, such as mechanical engineering, trains, power grids, equipment failures, etc. However, the summary of the system of fault KGs is relatively small. Therefore, this article provides a comprehensive tutorial and survey about the recent advances toward the construction of fault KG. Specifically, it will provide an overview of the fault KG and summarize the key techniques for building a KG to guide the construction of the KG in the fault domain. What’s more, it introduces some of the open source tools that can be used to build a KG process, enabling researchers and practitioners to quickly get started in this field. In addition, the article discusses the application of fault KG and the difficulties and challenges in constructing fault KG. Finally, the article looks forward to the future development of KG.

8 citations

Journal ArticleDOI
01 Aug 2021
TL;DR: In a Smart Factory, manufacturing systems act largely autonomously on the basis of continuously collected data and can autonomously adapt to events through the ability to intelligently analyze data and act as adaptive manufacturing systems that consider changes in production, the supply chain and customer requirements.
Abstract: Knowledge integration is well explained by the human–organization–technology (HOT) approach known from knowledge management. This approach contains the horizontal and vertical interaction and communication between employees, human-to-machine, but also machine-to-machine. Different organizational structures and processes are supported with the help of appropriate technologies and suitable data processing and integration techniques. In a Smart Factory, manufacturing systems act largely autonomously on the basis of continuously collected data. The technical design concerns the networking of machines, their connectivity and the interaction between human and machine as well as machine-to-machine. Within a Smart Factory, machines can be considered as intelligent manufacturing systems. Such manufacturing systems can autonomously adapt to events through the ability to intelligently analyze data and act as adaptive manufacturing systems that consider changes in production, the supply chain and customer requirements. Inter-connected physical devices, sensors, actuators, and controllers form the building block of the Smart Factory, which is called the Internet of Things (IoT). IoT uses different data processing solutions, such as cloud computing, fog computing, or edge computing, to fuse and process data. This is accomplished in an integrated and cross-device manner.

6 citations


Cites background from "Knowledge-Graph Based Multi-Target ..."

  • ...This is observed in the work of Qin et al. (2018), who suggest an approach for the anomaly detection problem based on a deep learning model that has been supported by a knowledge graph ....

    [...]

Journal ArticleDOI
TL;DR: Cheng et al. as discussed by the authors proposed hierarchical graphical reasoning (HGR), which utilizes the hierarchical structures of trains for train component detection, which contains multiple graphical reasoning branches, each of which is utilized to conduct graphical reasoning for one cluster of train components based on their sizes.
Abstract: Recently, deep learning-based approaches have achieved superior performance on object detection applications. However, object detection for industrial scenarios, where the objects may also have some structures and the structured patterns are normally presented in a hierarchical way, is not well investigated yet. In this work, we propose a novel deep learning-based method, hierarchical graphical reasoning (HGR), which utilizes the hierarchical structures of trains for train component detection. HGR contains multiple graphical reasoning branches, each of which is utilized to conduct graphical reasoning for one cluster of train components based on their sizes. In each branch, the visual appearances and structures of train components are considered jointly with our proposed novel densely connected dual-gated recurrent units (Dense-DGRUs). To the best of our knowledge, HGR is the first kind of framework that explores hierarchical structures among objects for object detection. We have collected a data set of 1130 images captured from moving trains, in which 17 334 train components are manually annotated with bounding boxes. Based on this data set, we carry out extensive experiments that have demonstrated our proposed HGR outperforms the existing state-of-the-art baselines significantly. The data set and the source code can be downloaded online at https://github.com/ChengZY/HGR .

2 citations

Proceedings ArticleDOI
26 May 2023
TL;DR: In this paper , a Fault Knowledge Graph (FKG) based on the knowledge graph (KG) in the field of faults is proposed for fault diagnosis, which can well analyze the relationship between various faults, achieve prediction, and promote development.
Abstract: In the field of equipment failure, a large number of Fault Tree Analysis(FTA) diagrams and Failure Mode, Effects and Criticality Analysis(FMECA) tables related to equipment failure have been precipitated, but they are not highly utilized for fault diagnosis. The Fault Knowledge Graph(FKG) is based on the Knowledge Graph(KG) in the field of faults, it can well analyze the relationship between various faults, achieve prediction, and promote development. Therefore, the FKG creation method based on FTA and FMECA is proposed for fault diagnosis. Due to the different types of data sources, we first define the data model of the FKG, adopt two different map creation methods, and then perform knowledge fusion to obtain the final FKG. A case is presented to demonstrate the effectiveness of our method.
References
More filters
Book ChapterDOI
05 Oct 2015
TL;DR: Neber et al. as discussed by the authors proposed a network and training strategy that relies on the strong use of data augmentation to use the available annotated samples more efficiently, which can be trained end-to-end from very few images and outperforms the prior best method (a sliding-window convolutional network) on the ISBI challenge for segmentation of neuronal structures in electron microscopic stacks.
Abstract: There is large consent that successful training of deep networks requires many thousand annotated training samples. In this paper, we present a network and training strategy that relies on the strong use of data augmentation to use the available annotated samples more efficiently. The architecture consists of a contracting path to capture context and a symmetric expanding path that enables precise localization. We show that such a network can be trained end-to-end from very few images and outperforms the prior best method (a sliding-window convolutional network) on the ISBI challenge for segmentation of neuronal structures in electron microscopic stacks. Using the same network trained on transmitted light microscopy images (phase contrast and DIC) we won the ISBI cell tracking challenge 2015 in these categories by a large margin. Moreover, the network is fast. Segmentation of a 512x512 image takes less than a second on a recent GPU. The full implementation (based on Caffe) and the trained networks are available at http://lmb.informatik.uni-freiburg.de/people/ronneber/u-net .

49,590 citations

Proceedings ArticleDOI
07 Jun 2015
TL;DR: Inception as mentioned in this paper is a deep convolutional neural network architecture that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14).
Abstract: We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.

40,257 citations

Proceedings ArticleDOI
07 Jun 2015
TL;DR: The key insight is to build “fully convolutional” networks that take input of arbitrary size and produce correspondingly-sized output with efficient inference and learning.
Abstract: Convolutional networks are powerful visual models that yield hierarchies of features. We show that convolutional networks by themselves, trained end-to-end, pixels-to-pixels, exceed the state-of-the-art in semantic segmentation. Our key insight is to build “fully convolutional” networks that take input of arbitrary size and produce correspondingly-sized output with efficient inference and learning. We define and detail the space of fully convolutional networks, explain their application to spatially dense prediction tasks, and draw connections to prior models. We adapt contemporary classification networks (AlexNet [20], the VGG net [31], and GoogLeNet [32]) into fully convolutional networks and transfer their learned representations by fine-tuning [3] to the segmentation task. We then define a skip architecture that combines semantic information from a deep, coarse layer with appearance information from a shallow, fine layer to produce accurate and detailed segmentations. Our fully convolutional network achieves state-of-the-art segmentation of PASCAL VOC (20% relative improvement to 62.2% mean IU on 2012), NYUDv2, and SIFT Flow, while inference takes less than one fifth of a second for a typical image.

28,225 citations

Posted Content
TL;DR: Faster R-CNN as discussed by the authors proposes a Region Proposal Network (RPN) to generate high-quality region proposals, which are used by Fast R-NN for detection.
Abstract: State-of-the-art object detection networks depend on region proposal algorithms to hypothesize object locations. Advances like SPPnet and Fast R-CNN have reduced the running time of these detection networks, exposing region proposal computation as a bottleneck. In this work, we introduce a Region Proposal Network (RPN) that shares full-image convolutional features with the detection network, thus enabling nearly cost-free region proposals. An RPN is a fully convolutional network that simultaneously predicts object bounds and objectness scores at each position. The RPN is trained end-to-end to generate high-quality region proposals, which are used by Fast R-CNN for detection. We further merge RPN and Fast R-CNN into a single network by sharing their convolutional features---using the recently popular terminology of neural networks with 'attention' mechanisms, the RPN component tells the unified network where to look. For the very deep VGG-16 model, our detection system has a frame rate of 5fps (including all steps) on a GPU, while achieving state-of-the-art object detection accuracy on PASCAL VOC 2007, 2012, and MS COCO datasets with only 300 proposals per image. In ILSVRC and COCO 2015 competitions, Faster R-CNN and RPN are the foundations of the 1st-place winning entries in several tracks. Code has been made publicly available.

23,183 citations

Posted Content
TL;DR: It is shown that such a network can be trained end-to-end from very few images and outperforms the prior best method (a sliding-window convolutional network) on the ISBI challenge for segmentation of neuronal structures in electron microscopic stacks.
Abstract: There is large consent that successful training of deep networks requires many thousand annotated training samples. In this paper, we present a network and training strategy that relies on the strong use of data augmentation to use the available annotated samples more efficiently. The architecture consists of a contracting path to capture context and a symmetric expanding path that enables precise localization. We show that such a network can be trained end-to-end from very few images and outperforms the prior best method (a sliding-window convolutional network) on the ISBI challenge for segmentation of neuronal structures in electron microscopic stacks. Using the same network trained on transmitted light microscopy images (phase contrast and DIC) we won the ISBI cell tracking challenge 2015 in these categories by a large margin. Moreover, the network is fast. Segmentation of a 512x512 image takes less than a second on a recent GPU. The full implementation (based on Caffe) and the trained networks are available at this http URL .

19,534 citations


"Knowledge-Graph Based Multi-Target ..." refers methods in this paper

  • ...This method was further improved in the U-Net neural network [5]-[7]....

    [...]