Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
4.5
2.7
Journals
Applied Sciences
Special Issues
Advances in Deep Learning Ⅱ
share announcement

Advances in Deep Learning Ⅱ

A special issue of Applied Sciences (ISSN 2076-3417). This special issue belongs to the section "Computing and Artificial Intelligence".

Deadline for manuscript submissions: closed (30 November 2020) | Viewed by 87116

Special Issue Editors

Prof. Dr. Wonjoon Kim

E-Mail Website
Guest Editor
Department of Industrial and Management Engineering, Sungkyul University, Seoul, Korea
Interests: deep learning; unstructured data analysis; human factors; explainable AI
Special Issues, Collections and Topics in MDPI journals
Prof. Dr. Jung Woo Baek

E-Mail Website
Guest Editor
Associate Professor, Department of Industrial Engineering, Chosun University, Gwangju, Korea
Interests: probabilistic modeling; queuing analysis; Markov process; operation research; reinforcement learning
Prof. Dr. Youngdoo Son

E-Mail Website
Guest Editor
Assistant Professor, Department of Industrial and Systems Engineering, Dongguk University - Seoul, Seoul, Korea
Interests: deep learning; meta learning; explainable AI; open-world classification; machine learning methods

Special Issue Information

Dear Colleagues,

The concept of deep learning has been applied to numerous research areas, such as prediction, classification, image/speech recognition, vision, etc., and has drastically surpassed traditional methodologies. The main difference between other methods and deep learning lies in providing a computational model for multiprocessing neural network layers by learning and representing data at multiple levels. Therefore, deep learning is easy to implicitly understand the complex structure of large data sets.

Therefore, deep learning models are able to give intuitions to understand the complex structures of large data sets. Recently, deep learning methods have actively been extended to other parts of machine learning, including reinforcement learning and transfer/meta learning, while standard deep learning methods such as RNN (recurrent neural network) and CNN (convolutional neural network) algorithms have also been extensively studied and applied to diverse industrial fields. The aim of our Special Issue is to examine the latest theoretical and practical applications of deep learning to various fields. In this connection, research related to the application of the newest deep learning techniques to image processing and data analysis can be used to improve the models derived from previous studies.

The purpose of our Special Issue is to contribute to the demonstration of new algorithms and application domains of deep learning to solve problems in various research areas. Eventually, we are to promote research and development of deep learning for multimodal data, by publishing high-quality research articles and reviews/tutorials in this rapidly growing interdisciplinary field.

Dr. Wonjoon Kim
Prof. Dr. Youngdoo Son
Prof. Dr. Jung Woo Baek
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Applied Sciences is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • deep learning algorithms
  • deep reinforcement learning
  • deep learning theory
  • deep learning architectures
  • adversarial examples
  • deep generative models
  • multitask, transfer, and meta learning
  • Bayesian methods
  • causal inferences
  • explainable and interpretable AI
  • representation learning
  • constrained optimization
  • computer vision
  • natural language processing
  • application in any other field using deep learning methods

Published Papers (20 papers)

Download All Papers
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:

Research

19 pages, 3335 KiB  
Article
Development of Defect Detection AI Model for Wire + Arc Additive Manufacturing Using High Dynamic Range Images
by Chaekyo Lee, Gijeong Seo, Duck Bong Kim, Minjae Kim and Jong-Ho Shin
Appl. Sci. 2021, 11(16), 7541; https://doi.org/10.3390/app11167541 - 17 Aug 2021
Cited by 23 | Viewed by 3573
Abstract
Wire + arc additive manufacturing (WAAM) utilizes a welding arc as a heat source and a metal wire as a feedstock. In recent years, WAAM has attracted significant attention in the manufacturing industry owing to its advantages: (1) high deposition rate, (2) low [...] Read more.
Wire + arc additive manufacturing (WAAM) utilizes a welding arc as a heat source and a metal wire as a feedstock. In recent years, WAAM has attracted significant attention in the manufacturing industry owing to its advantages: (1) high deposition rate, (2) low system setup cost, (3) wide diversity of wire materials, and (4) sustainability for constructing large-sized metal structures. However, owing to the complexity of arc welding in WAAM, more research efforts are required to improve its process repeatability and advance part qualification. This study proposes a methodology to detect defects of the arch welding process in WAAM using images acquired by a high dynamic range camera. The gathered images are preprocessed to emphasize features and used for an artificial intelligence model to classify normal and abnormal statuses of arc welding in WAAM. Owing to the shortage of image datasets for defects, transfer learning technology is adopted. In addition, to understand and check the basis of the model’s feature learning, a gradient-weighted class activation mapping algorithm is applied to select a model that has the correct judgment criteria. Experimental results show that the detection accuracy of the metal transfer region-of-interest (RoI) reached 99%, whereas that of the weld-pool and bead RoI was 96%. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

18 pages, 19002 KiB  
Article
Frequency-Based Haze and Rain Removal Network (FHRR-Net) with Deep Convolutional Encoder-Decoder
by Dong Hwan Kim, Woo Jin Ahn, Myo Taeg Lim, Tae Koo Kang and Dong Won Kim
Appl. Sci. 2021, 11(6), 2873; https://doi.org/10.3390/app11062873 - 23 Mar 2021
Cited by 4 | Viewed by 2892
Abstract
Removing haze or rain is one of the difficult problems in computer vision applications. On real-world road images, haze and rain often occur together, but traditional methods cannot solve this imaging problem. To address rain and haze problems simultaneously, we present a robust [...] Read more.
Removing haze or rain is one of the difficult problems in computer vision applications. On real-world road images, haze and rain often occur together, but traditional methods cannot solve this imaging problem. To address rain and haze problems simultaneously, we present a robust network-based framework consisting of three steps: image decomposition using guided filters, a frequency-based haze and rain removal network (FHRR-Net), and image restoration based on an atmospheric scattering model using predicted transmission maps and predicted rain-removed images. We demonstrate FHRR-Net’s capabilities with synthesized and real-world road images. Experimental results show that our trained framework has superior performance on synthesized and real-world road test images compared with state-of-the-art methods. We use PSNR (peak signal-to-noise) and SSIM (structural similarity index) indicators to evaluate our model quantitatively, showing that our methods have the highest PSNR and SSIM values. Furthermore, we demonstrate through experiments that our method is useful in real-world vision applications. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

23 pages, 3292 KiB  
Article
Augmented CWT Features for Deep Learning-Based Indoor Localization Using WiFi RSSI Data
by Paul Ssekidde, Odongo Steven Eyobu, Dong Seog Han and Tonny J. Oyana
Appl. Sci. 2021, 11(4), 1806; https://doi.org/10.3390/app11041806 - 18 Feb 2021
Cited by 34 | Viewed by 4135
Abstract
Localization is one of the current challenges in indoor navigation research. The conventional global positioning system (GPS) is affected by weak signal strengths due to high levels of signal interference and fading in indoor environments. Therefore, new positioning solutions tailored for indoor environments [...] Read more.
Localization is one of the current challenges in indoor navigation research. The conventional global positioning system (GPS) is affected by weak signal strengths due to high levels of signal interference and fading in indoor environments. Therefore, new positioning solutions tailored for indoor environments need to be developed. In this paper, we propose a deep learning approach for indoor localization. However, the performance of a deep learning system depends on the quality of the feature representation. This paper introduces two novel feature set extractions based on the continuous wavelet transforms (CWT) of the received signal strength indicators’ (RSSI) data. The two novel CWT feature sets were augmented with additive white Gaussian noise. The first feature set is CWT image-based, and the second is composed of the CWT PSD numerical data that were dimensionally equalized using principal component analysis (PCA). These proposed image and numerical data feature sets were both evaluated using CNN and ANN models with the goal of identifying the room that the human subject was in and estimating the precise location of the human subject in that particular room. Extensive experiments were conducted to generate the proposed augmented CWT feature set and numerical CWT PSD feature set using two analyzing functions, namely, Morlet and Morse. For validation purposes, the performance of the two proposed feature sets were compared with each other and other existing feature set formulations. The accuracy, precision and recall results show that the proposed feature sets performed better than the conventional feature sets used to validate the study. Similarly, the mean localization error generated by the proposed feature set predictions was less than those of the conventional feature sets used in indoor localization. More particularly, the proposed augmented CWT-image feature set outperformed the augmented CWT-PSD numerical feature set. The results also show that the Morse-based feature sets trained with CNN produced the best indoor positioning results compared to all Morlet and ANN-based feature set formulations. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

16 pages, 7840 KiB  
Article
Seg2pix: Few Shot Training Line Art Colorization with Segmented Image Data
by Chang Wook Seo and Yongduek Seo
Appl. Sci. 2021, 11(4), 1464; https://doi.org/10.3390/app11041464 - 05 Feb 2021
Cited by 10 | Viewed by 4172
Abstract
There are various challenging issues in automating line art colorization. In this paper, we propose a GAN approach incorporating semantic segmentation image data. Our GAN-based method, named Seg2pix, can automatically generate high quality colorized images, aiming at computerizing one of the most tedious [...] Read more.
There are various challenging issues in automating line art colorization. In this paper, we propose a GAN approach incorporating semantic segmentation image data. Our GAN-based method, named Seg2pix, can automatically generate high quality colorized images, aiming at computerizing one of the most tedious and repetitive jobs performed by coloring workers in the webtoon industry. The network structure of Seg2pix is mostly a modification of the architecture of Pix2pix, which is a convolution-based generative adversarial network for image-to-image translation. Through this method, we can generate high quality colorized images of a particular character with only a few training data. Seg2pix is designed to reproduce a segmented image, which becomes the suggestion data for line art colorization. The segmented image is automatically generated through a generative network with a line art image and a segmentation ground truth. In the next step, this generative network creates a colorized image from the line art and segmented image, which is generated from the former step of the generative network. To summarize, only one line art image is required for testing the generative model, and an original colorized image and segmented image are additionally required as the ground truth for training the model. These generations of the segmented image and colorized image proceed by an end-to-end method sharing the same loss functions. By using this method, we produce better qualitative results for automatic colorization of a particular character’s line art. This improvement can also be measured by quantitative results with Learned Perceptual Image Patch Similarity (LPIPS) comparison. We believe this may help artists exercise their creative expertise mainly in the area where computerization is not yet capable. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

12 pages, 6477 KiB  
Article
Semantic Scene Graph Generation Using RDF Model and Deep Learning
by Seongyong Kim, Tae Hyeon Jeon, Ilsun Rhiu, Jinhyun Ahn and Dong-Hyuk Im
Appl. Sci. 2021, 11(2), 826; https://doi.org/10.3390/app11020826 - 17 Jan 2021
Cited by 3 | Viewed by 2688
Abstract
Over the last several years, in parallel with the general global advancement in mobile technology and a rise in social media network content consumption, multimedia content production and reproduction has increased exponentially. Therefore, enabled by the rapid recent advancements in deep learning technology, [...] Read more.
Over the last several years, in parallel with the general global advancement in mobile technology and a rise in social media network content consumption, multimedia content production and reproduction has increased exponentially. Therefore, enabled by the rapid recent advancements in deep learning technology, research on scene graph generation is being actively conducted to more efficiently search for and classify images desired by users within a large amount of content. This approach lets users accurately find images they are searching for by expressing meaningful information on image content as nodes and edges of a graph. In this study, we propose a scene graph generation method based on using the Resource Description Framework (RDF) model to clarify semantic relations. Furthermore, we also use convolutional neural network (CNN) and recurrent neural network (RNN) deep learning models to generate a scene graph expressed in a controlled vocabulary of the RDF model to understand the relations between image object tags. Finally, we experimentally demonstrate through testing that our proposed technique can express semantic content more effectively than existing approaches. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

12 pages, 3283 KiB  
Article
Korean Historical Documents Analysis with Improved Dynamic Word Embedding
by KyoHoon Jin, JeongA Wi, KyeongPil Kang and YoungBin Kim
Appl. Sci. 2020, 10(21), 7939; https://doi.org/10.3390/app10217939 - 09 Nov 2020
Cited by 2 | Viewed by 2576
Abstract
Historical documents refer to records or books that provide textual information about the thoughts and consciousness of past civilisations, and therefore, they have historical significance. These documents are used as key sources for historical studies as they provide information over several historical periods. [...] Read more.
Historical documents refer to records or books that provide textual information about the thoughts and consciousness of past civilisations, and therefore, they have historical significance. These documents are used as key sources for historical studies as they provide information over several historical periods. Many studies have analysed various historical documents using deep learning; however, studies that employ changes in information over time are lacking. In this study, we propose a deep-learning approach using improved dynamic word embedding to determine the characteristics of 27 kings mentioned in the Annals of the Joseon Dynasty, which contains a record of 500 years. The characteristics of words for each king were quantitated based on dynamic word embedding; further, this information was applied to named entity recognition and neural machine translation.In experiments, we confirmed that the method we proposed showed better performance than other methods. In the named entity recognition task, the F1-score was 0.68; in the neural machine translation task, the BLEU4 score was 0.34. We demonstrated that this approach can be used to extract information about diplomatic relationships with neighbouring countries and the economic conditions of the Joseon Dynasty. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

10 pages, 3045 KiB  
Article
Expert System for Mandibular Condyle Detection and Osteoarthritis Classification in Panoramic Imaging Using R-CNN and CNN
by Donghyun Kim, Eunhye Choi, Ho Gul Jeong, Joonho Chang and Sekyoung Youm
Appl. Sci. 2020, 10(21), 7464; https://doi.org/10.3390/app10217464 - 23 Oct 2020
Cited by 16 | Viewed by 9316
Abstract
Temporomandibular joint osteoarthritis (TMJ OA) is a degenerative condition of the TMJ led by a pathological tissue response of the joint under mechanical loading. It is characterized by the progressive destruction of the internal surfaces of the joint, which can result in debilitating [...] Read more.
Temporomandibular joint osteoarthritis (TMJ OA) is a degenerative condition of the TMJ led by a pathological tissue response of the joint under mechanical loading. It is characterized by the progressive destruction of the internal surfaces of the joint, which can result in debilitating pain and joint noise. Panoramic imaging can be used as a basic screening tool with thorough clinical examination in diagnosing TMJ OA. This paper proposes an algorithm that can extract the condylar region and determine its abnormality by using convolutional neural networks (CNNs) and Faster region-based CNNs (R-CNNs). Panoramic images are collected retrospectively and 1000 images are classified into three categories—normal, abnormal, and unreadable—by a dentist or orofacial pain specialist. Labels indicating whether the condyle is detected and its location enabled more clearly recognizable panoramic images. The uneven proportion of normal to abnormal data is adjusted by duplicating and rotating the images. An R-CNN model and a Visual Geometry Group-16 (VGG16) model are used for learning and condyle discrimination, respectively. To prevent overfitting, the images are rotated ±10° and shifted by 10%. The average precision of condyle detection using an R-CNN at intersection over union (IoU) >0.5 is 99.4% (right side) and 100% (left side). The sensitivity, specificity, and accuracy of the TMJ OA classification algorithm using a CNN are 0.54, 0.94, and 0.84, respectively. The findings demonstrate that classifying panoramic images through CNNs is possible. It is expected that artificial intelligence will be more actively applied to analyze panoramic X-ray images in the future. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

19 pages, 7668 KiB  
Article
Semantics-Preserving RDB2RDF Data Transformation Using Hierarchical Direct Mapping
by Hee-Gook Jun and Dong-Hyuk Im
Appl. Sci. 2020, 10(20), 7070; https://doi.org/10.3390/app10207070 - 12 Oct 2020
Cited by 2 | Viewed by 2319
Abstract
Direct mapping is an automatic transformation method used to generate resource description framework (RDF) data from relational data. In the field of direct mapping, semantics preservation is critical to ensure that the mapping method outputs RDF data without information loss or incorrect semantic [...] Read more.
Direct mapping is an automatic transformation method used to generate resource description framework (RDF) data from relational data. In the field of direct mapping, semantics preservation is critical to ensure that the mapping method outputs RDF data without information loss or incorrect semantic data generation. However, existing direct-mapping methods have problems that prevent semantics preservation in specific cases. For this reason, a mapping method is developed to perform a semantics-preserving transformation of relational databases (RDB) into RDF data without semantic information loss and to reduce the volume of incorrect RDF data. This research reviews cases that do not generate semantics-preserving results, and the corresponding problems into categories are arranged. This paper defines lemmas that represent the features of RDF data transformation to resolve those problems. Based on the lemmas, this work develops a hierarchical direct-mapping method to strictly abide by the definition of semantics preservation and to prevent semantic information loss, reducing the volume of incorrect RDF data generated. Experiments demonstrate the capability of the proposed method to perform semantics-preserving RDB2RDF data transformation, generating semantically accurate results. This work impacts future studies, which should involve the development of synchronization methods to achieve RDF data consistency when original RDB data are modified. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

13 pages, 1298 KiB  
Article
VB-Net: Voxel-Based Broad Learning Network for 3D Object Classification
by Zishu Liu, Wei Song, Yifei Tian, Sumi Ji, Yunsick Sung, Long Wen, Tao Zhang, Liangliang Song and Amanda Gozho
Appl. Sci. 2020, 10(19), 6735; https://doi.org/10.3390/app10196735 - 26 Sep 2020
Cited by 18 | Viewed by 3359
Abstract
Point clouds have been widely used in three-dimensional (3D) object classification tasks, i.e., people recognition in unmanned ground vehicles. However, the irregular data format of point clouds and the large number of parameters in deep learning networks affect the performance of object classification. [...] Read more.
Point clouds have been widely used in three-dimensional (3D) object classification tasks, i.e., people recognition in unmanned ground vehicles. However, the irregular data format of point clouds and the large number of parameters in deep learning networks affect the performance of object classification. This paper develops a 3D object classification system using a broad learning system (BLS) with a feature extractor called VB-Net. First, raw point clouds are voxelized into voxels. Through this step, irregular point clouds are converted into regular voxels which are easily processed by the feature extractor. Then, a pre-trained VoxNet is employed as a feature extractor to extract features from voxels. Finally, those features are used for object classification by the applied BLS. The proposed system is tested on the ModelNet40 dataset and ModelNet10 dataset. The average recognition accuracy was 83.99% and 90.08%, respectively. Compared to deep learning networks, the time consumption of the proposed system is significantly decreased. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

12 pages, 4011 KiB  
Article
Automatic Tooth Detection and Numbering Using a Combination of a CNN and Heuristic Algorithm
by Changgyun Kim, Donghyun Kim, HoGul Jeong, Suk-Ja Yoon and Sekyoung Youm
Appl. Sci. 2020, 10(16), 5624; https://doi.org/10.3390/app10165624 - 13 Aug 2020
Cited by 32 | Viewed by 6136
Abstract
Dental panoramic radiography (DPR) is a method commonly used in dentistry for patient diagnosis. This study presents a new technique that combines a regional convolutional neural network (RCNN), Single Shot Multibox Detector, and heuristic methods to detect and number the teeth and implants [...] Read more.
Dental panoramic radiography (DPR) is a method commonly used in dentistry for patient diagnosis. This study presents a new technique that combines a regional convolutional neural network (RCNN), Single Shot Multibox Detector, and heuristic methods to detect and number the teeth and implants with only fixtures in a DPR image. This technology is highly significant in providing statistical information and personal identification based on DPR and separating the images of individual teeth, which serve as basic data for various DPR-based AI algorithms. As a result, the mAP(@IOU = 0.5) of the tooth, implant fixture, and crown detection using the RCNN algorithm were obtained at rates of 96.7%, 45.1%, and 60.9%, respectively. Further, the sensitivity, specificity, and accuracy of the tooth numbering algorithm using a convolutional neural network and heuristics were 84.2%, 75.5%, and 84.5%, respectively. Techniques to analyze DPR images, including implants and bridges, were developed, enabling the possibility of applying AI to orthodontic or implant DPR images of patients. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

13 pages, 499 KiB  
Article
Missing Value Imputation in Stature Estimation by Learning Algorithms Using Anthropometric Data: A Comparative Study
by Youngdoo Son and Wonjoon Kim
Appl. Sci. 2020, 10(14), 5020; https://doi.org/10.3390/app10145020 - 21 Jul 2020
Cited by 7 | Viewed by 2371
Abstract
Estimating stature is essential in the process of personal identification. Because it is difficult to find human remains intact at crime scenes and disaster sites, for instance, methods are needed for estimating stature based on different body parts. For instance, the upper and [...] Read more.
Estimating stature is essential in the process of personal identification. Because it is difficult to find human remains intact at crime scenes and disaster sites, for instance, methods are needed for estimating stature based on different body parts. For instance, the upper and lower limbs may vary depending on ancestry and sex, and it is of great importance to design adequate methodology for incorporating these in estimating stature. In addition, it is necessary to use machine learning rather than simple linear regression to improve the accuracy of stature estimation. In this study, the accuracy of statures estimated based on anthropometric data was compared using three imputation methods. In addition, by comparing the accuracy among linear and nonlinear classification methods, the best method was derived for estimating stature based on anthropometric data. For both sexes, multiple imputation was superior when the missing data ratio was low, and mean imputation performed well when the ratio was high. The support vector machine recorded the highest accuracy in all ratios of missing data. The findings of this study showed appropriate imputation methods for estimating stature with missing anthropometric data. In particular, the machine learning algorithms can be effectively used for estimating stature in humans. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

25 pages, 2325 KiB  
Article
Time-Series Classification Based on Fusion Features of Sequence and Visualization
by Baoquan Wang, Tonghai Jiang, Xi Zhou, Bo Ma, Fan Zhao and Yi Wang
Appl. Sci. 2020, 10(12), 4124; https://doi.org/10.3390/app10124124 - 15 Jun 2020
Cited by 8 | Viewed by 4762
Abstract
For the task of time-series data classification (TSC), some methods directly classify raw time-series (TS) data. However, certain sequence features are not evident in the time domain and the human brain can extract visual features based on visualization to classify data. Therefore, some [...] Read more.
For the task of time-series data classification (TSC), some methods directly classify raw time-series (TS) data. However, certain sequence features are not evident in the time domain and the human brain can extract visual features based on visualization to classify data. Therefore, some researchers have converted TS data to image data and used image processing methods for TSC. While human perceptionconsists of a combination of human senses from different aspects, existing methods only use sequence features or visualization features. Therefore, this paper proposes a framework for TSC based on fusion features (TSC-FF) of sequence features extracted from raw TS and visualization features extracted from Area Graphs converted from TS. Deep learning methods have been proven to be useful tools for automatically learning features from data; therefore, we use long short-term memory with an attention mechanism (LSTM-A) to learn sequence features and a convolutional neural network with an attention mechanism (CNN-A) for visualization features, in order to imitate the human brain. In addition, we use the simplest visualization method of Area Graph for visualization features extraction, avoiding loss of information and additional computational cost. This article aims to prove that using deep neural networks to learn features from different aspects and fusing them can replace complex, artificially constructed features, as well as remove the bias due to manually designed features, in order to avoid the limitations of domain knowledge. Experiments on several open data sets show that the framework achieves promising results, compared with other methods. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

17 pages, 1195 KiB  
Article
FedOpt: Towards Communication Efficiency and Privacy Preservation in Federated Learning
by Muhammad Asad, Ahmed Moustafa and Takayuki Ito
Appl. Sci. 2020, 10(8), 2864; https://doi.org/10.3390/app10082864 - 21 Apr 2020
Cited by 69 | Viewed by 8339
Abstract
Artificial Intelligence (AI) has been applied to solve various challenges of real-world problems in recent years. However, the emergence of new AI technologies has brought several problems, especially with regard to communication efficiency, security threats and privacy violations. Towards this end, Federated Learning [...] Read more.
Artificial Intelligence (AI) has been applied to solve various challenges of real-world problems in recent years. However, the emergence of new AI technologies has brought several problems, especially with regard to communication efficiency, security threats and privacy violations. Towards this end, Federated Learning (FL) has received widespread attention due to its ability to facilitate the collaborative training of local learning models without compromising the privacy of data. However, recent studies have shown that FL still consumes considerable amounts of communication resources. These communication resources are vital for updating the learning models. In addition, the privacy of data could still be compromised once sharing the parameters of the local learning models in order to update the global model. Towards this end, we propose a new approach, namely, Federated Optimisation (FedOpt) in order to promote communication efficiency and privacy preservation in FL. In order to implement FedOpt, we design a novel compression algorithm, namely, Sparse Compression Algorithm (SCA) for efficient communication, and then integrate the additively homomorphic encryption with differential privacy to prevent data from being leaked. Thus, the proposed FedOpt smoothly trade-offs communication efficiency and privacy preservation in order to adopt the learning task. The experimental results demonstrate that FedOpt outperforms the state-of-the-art FL approaches. In particular, we consider three different evaluation criteria; model accuracy, communication efficiency and computation overhead. Then, we compare the proposed FedOpt with the baseline configurations and the state-of-the-art approaches, i.e., Federated Averaging (FedAvg) and the paillier-encryption based privacy-preserving deep learning (PPDL) on all these three evaluation criteria. The experimental results show that FedOpt is able to converge within fewer training epochs and a smaller privacy budget. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

15 pages, 1034 KiB  
Article
Go Wider: An Efficient Neural Network for Point Cloud Analysis via Group Convolutions
by Can Chen, Luca Zanotti Fragonara and Antonios Tsourdos
Appl. Sci. 2020, 10(7), 2391; https://doi.org/10.3390/app10072391 - 01 Apr 2020
Cited by 6 | Viewed by 2827
Abstract
In order to achieve a better performance for point cloud analysis, many researchers apply deep neural networks using stacked Multi-Layer-Perceptron (MLP) convolutions over an irregular point cloud. However, applying these dense MLP convolutions over a large amount of points (e.g., autonomous driving application) [...] Read more.
In order to achieve a better performance for point cloud analysis, many researchers apply deep neural networks using stacked Multi-Layer-Perceptron (MLP) convolutions over an irregular point cloud. However, applying these dense MLP convolutions over a large amount of points (e.g., autonomous driving application) leads to limitations due to the computation and memory capabilities. To achieve higher performances but decrease the computational complexity, we propose a deep-wide neural network, named ShufflePointNet, which can exploit fine-grained local features, but also reduce redundancies using group convolution and channel shuffle operation. Unlike conventional operations that directly apply MLPs on the high-dimensional features of a point cloud, our model goes “wider” by splitting features into groups with smaller depth in advance, having the respective MLP computations applied only to a single group, which can significantly reduce complexity and computation. At the same time, we allow communication between groups by shuffling the feature channel to capture fine-grained features. We further discuss the multi-branch method for wider neural networks being also beneficial to feature extraction for point clouds. We present extensive experiments for shape classification tasks on a ModelNet40 dataset and semantic segmentation task on large scale datasets ShapeNet part, S3DIS and KITTI. Finally, we carry out an ablation study and compare our model to other state-of-the-art algorithms to show its efficiency in terms of complexity and accuracy. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

16 pages, 1829 KiB  
Article
DUKMSVM: A Framework of Deep Uniform Kernel Mapping Support Vector Machine for Short Text Classification
by Zhaoying Liu, Haipeng Kan, Ting Zhang and Yujian Li
Appl. Sci. 2020, 10(7), 2348; https://doi.org/10.3390/app10072348 - 30 Mar 2020
Cited by 10 | Viewed by 2078
Abstract
This paper mainly deals with the problem of short text classification. There are two main contributions. Firstly, we introduce a framework of deep uniform kernel mapping support vector machine (DUKMSVM). The significant merit of this framework is that by expressing the [...] Read more.
This paper mainly deals with the problem of short text classification. There are two main contributions. Firstly, we introduce a framework of deep uniform kernel mapping support vector machine (DUKMSVM). The significant merit of this framework is that by expressing the kernel mapping function explicitly with a deep neural network, it is in essence an explicit kernel mapping instead of the traditional kernel function, and it allows better flexibility in dealing with various applications by applying different neural network structures. Secondly, to validate the effectiveness of this framework and to improve the performance of short text classification, we explicitly express the kernel mapping using bidirectional recurrent neural network (BRNN), and propose a deep bidirectional recurrent kernel mapping support vector machine (DRKMSVM) for short text classification. Experimental results on five public short text classification datasets indicate that in terms of classification accuracy, precision, recall rate and F1-score, the DRKMSVM achieves the best performance with the average values of accuracy, precision, recall rate, and F1-score of 87.23%, 86.99%, 86.13% and 86.51% respectively compared to traditional SVM, convolutional neural network (CNN), Naive Bayes (NB), and Deep Neural Mapping Support Vector Machine (DNMSVM) which applies multi-layer perceptron for kernel mapping. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

16 pages, 978 KiB  
Article
Adaptive Real-Time Offloading Decision-Making for Mobile Edges: Deep Reinforcement Learning Framework and Simulation Results
by Soohyun Park, Dohyun Kwon, Joongheon Kim, Youn Kyu Lee and Sungrae Cho
Appl. Sci. 2020, 10(5), 1663; https://doi.org/10.3390/app10051663 - 01 Mar 2020
Cited by 12 | Viewed by 4161
Abstract
This paper proposes a novel dynamic offloading decision method which is inspired by deep reinforcement learning (DRL). In order to realize real-time communications in mobile edge computing systems, an efficient task offloading algorithm is required. When the decision of actions (offloading enabled, i.e., [...] Read more.
This paper proposes a novel dynamic offloading decision method which is inspired by deep reinforcement learning (DRL). In order to realize real-time communications in mobile edge computing systems, an efficient task offloading algorithm is required. When the decision of actions (offloading enabled, i.e., computing in clouds or offloading disabled, i.e., computing in local edges) is made by the proposed DRL-based dynamic algorithm in each unit time, it is required to consider real-time/seamless data transmission and energy-efficiency in mobile edge devices. Therefore, our proposed dynamic offloading decision algorithm is designed for the joint optimization of delay and energy-efficient communications based on DRL framework. According to the performance evaluation via data-intensive simulations, this paper verifies that the proposed dynamic algorithm achieves desired performance. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

24 pages, 1998 KiB  
Article
Forecasting Daily Temperatures with Different Time Interval Data Using Deep Neural Networks
by Sungjae Lee, Yung-Seop Lee and Youngdoo Son
Appl. Sci. 2020, 10(5), 1609; https://doi.org/10.3390/app10051609 - 28 Feb 2020
Cited by 30 | Viewed by 6357
Abstract
Temperature forecasting has been a consistent research topic owing to its significant effect on daily lives and various industries. However, it is an ever-challenging task because temperature is affected by various climate factors. Research on temperature forecasting has taken one of two directions: [...] Read more.
Temperature forecasting has been a consistent research topic owing to its significant effect on daily lives and various industries. However, it is an ever-challenging task because temperature is affected by various climate factors. Research on temperature forecasting has taken one of two directions: time-series analysis and machine learning algorithms. Recently, a large amount of high-frequent climate data have been well-stored and become available. In this study, we apply three types of neural networks, multilayer perceptron, recurrent, and convolutional, to daily average, minimum, and maximum temperature forecasting with higher-frequency input features than researchers used in previous studies. Applying these neural networks to the observed data from three locations with different climate characteristics, we show that prediction performance with highly frequent hourly input data is better than forecasting performance with less-frequent daily inputs. We observe that a convolutional neural network, which has been mostly employed for processing satellite images rather than numeric weather data for temperature forecasting, outperforms the other models. In addition, we combine state of the art weather forecasting techniques with the convolutional neural network and evaluate their effects on the temperature forecasting performances. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

20 pages, 5892 KiB  
Article
Flexible Fashion Product Retrieval Using Multimodality-Based Deep Learning
by Yeonsik Jo, Jehyeon Wi, Minseok Kim and Jae Yeol Lee
Appl. Sci. 2020, 10(5), 1569; https://doi.org/10.3390/app10051569 - 25 Feb 2020
Cited by 5 | Viewed by 4548
Abstract
Typically, fashion product searching in online shopping malls uses meta-information of the product. However, the use of meta-information is not guaranteed to ensure customer satisfaction, because of inherent limitations on the inaccuracy of input meta-information, imbalance of categories, and misclassification of apparel images. [...] Read more.
Typically, fashion product searching in online shopping malls uses meta-information of the product. However, the use of meta-information is not guaranteed to ensure customer satisfaction, because of inherent limitations on the inaccuracy of input meta-information, imbalance of categories, and misclassification of apparel images. These limitations prevent the shopping mall from providing a user-desired product retrieval. This paper proposes a new fashion product search method using multimodality-based deep learning, which can support more flexible and efficient retrieval by combining faceted queries and fashion image-based features. A deep convolutional neural network (CNN) generates a unique feature vector of the image, and the query input by the user is vectorized through a long short-term memory (LSTM)-based recurrent neural network (RNN). Then, the semantic similarity between the query vector and the product image vector is calculated to obtain the best match. Three different forms of the faceted query are supported. We perform quantitative and qualitative analyses to prove the effectiveness and originality of the proposed approach. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

16 pages, 5439 KiB  
Article
DERN: Deep Ensemble Learning Model for Short- and Long-Term Prediction of Baltic Dry Index
by Imam Mustafa Kamal, Hyerim Bae, Sim Sunghyun and Heesung Yun
Appl. Sci. 2020, 10(4), 1504; https://doi.org/10.3390/app10041504 - 22 Feb 2020
Cited by 33 | Viewed by 4378
Abstract
The Baltic Dry Index (BDI) is a commonly utilized indicator of global shipping and trade activity. It influences stakeholders’ and ship-owners’ decisions respecting investments, chartering, operational plans, and export and import activities. Accurate prediction of the BDI is very challenging due to its [...] Read more.
The Baltic Dry Index (BDI) is a commonly utilized indicator of global shipping and trade activity. It influences stakeholders’ and ship-owners’ decisions respecting investments, chartering, operational plans, and export and import activities. Accurate prediction of the BDI is very challenging due to its volatility, non-stationarity, and complexity. To help stakeholders and ship-owners make sound short- and long-term maritime business decisions and avoid market risk, we performed short- and long-term predictions of BDI using an ensemble deep-learning approach. In this study, we propose to apply recurrent neural network models for BDI prediction. The state-of-the-art of sequential deep-learning models such as RNN, LSTM, and GRU are employed to predict one- and multi-step-ahead BDI values. In order to increase the accuracy, we assemble the models. In experiments, we compared our results with those of traditional methods such as ARIMA and MLP. The results showed that our proposed method outperforms ARIMA, MLP, RNN, LSTM, and GRU in both short- and long-term prediction of BDI. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

9 pages, 3408 KiB  
Article
Detecting and Localizing Dents on Vehicle Bodies Using Region-Based Convolutional Neural Network
by Sung Hyun Park, Amir Tjolleng, Joonho Chang, Myeongsup Cha, Jongcheol Park and Kihyo Jung
Appl. Sci. 2020, 10(4), 1250; https://doi.org/10.3390/app10041250 - 13 Feb 2020
Cited by 8 | Viewed by 3624
Abstract
Detection and localization of the dents on a vehicle body that occurs during manufacturing is critical to achieve the appearance quality of a new vehicle. This study proposes a region-based convolutional neural network (R-CNN) to detect and localize dents for a vehicle body [...] Read more.
Detection and localization of the dents on a vehicle body that occurs during manufacturing is critical to achieve the appearance quality of a new vehicle. This study proposes a region-based convolutional neural network (R-CNN) to detect and localize dents for a vehicle body inspection. For a better feature extraction, this study employed a lighting system, which can highlight dents on an image by projecting the Mach bands (bright-dark stripes). The R-CNN was trained using the highlighted images by the Mach bands, and heat-maps were prepared with the classification scores estimated from the R-CNN to localize dents. This study applied the proposed R-CNN to the inspection of dents on the surface of a car body and quantitatively analyzed its performances. The detection accuracy of the dents was 98.5% for the testing data set, and mean absolute error between the actual dents and estimated dents were 13.7 pixels, which were close to one another. The proposed R-CNN could be applied to detect and localize surface dents during the manufacture of vehicle bodies in the automobile industry. Full article
(This article belongs to the Special Issue Advances in Deep Learning Ⅱ)
►▼ Show Figures

Figure 1

Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-20
Back to TopTop