scholarly journals A novel deep neural network for hidden target detection in images

2021 ◽  
Author(s):  
Rabeb Hendaoui ◽  
◽  
Vasif Nabiyev ◽  
Keyword(s):  
Neural Network ◽  
Target Detection ◽  
Deep Neural Network ◽  
Target Identification ◽  
Input Image ◽  
Second Phase ◽  
Natural Environments ◽  
Significant Similarity ◽  
Target Segmentation ◽  
Two Phases

The significant similarity between the hidden target and the background makes it difficult to find camouflaged people, such as warriors in warfare, or even camouflaged objects in natural environments. Hence, it is hard to ascertain these concealed targets. To address this issue, a novel deep neural network is proposed in this paper that produces an estimated mask within the hidden target for an input image. Our approach consists of two phases: hidden target segmentation and hidden target identification. For the first phase, we propose the Multilevel Attention Network (MA-Net), which generates the camouflaged target mask based on a Multi-Attention Module (MAM) that helps distinguish the hidden people from the background. Later on, the concealed target will be highlighted in the second phase. Experimental results on the camouflaged people dataset demonstrate that our proposed method can achieve state-of-the-art performance for hidden target detection.

scholarly journals Optimized Multi-Model Biometric Based Human Authentication using Deep Neural Network

2019 ◽  
Vol 8 (3S3) ◽  
pp. 280-290
Keyword(s):  
Neural Network ◽  
Deep Neural Network ◽  
Histogram Equalization ◽  
Extraction Process ◽  
Input Image ◽  
Suggested Approach ◽  
Biometric Systems ◽  
Recognition Systems ◽  
The Right ◽  
Facial Identification

Biometrics provides greater security and usability than conventional personal authentication methods. Fingerprints, facial identification systems and voice recognition systems are the features that biometric systems can use. To improve biometric authentication, the proposed method considered that the input image is iris and fingerprint; at first, pre-processing is performed through histogram equalization for all image inputs to enhance the image quality. Then the extraction process of the feature will be performed. The suggested method uses modified Local Binary Pattern (MLBP), GLCM with orientation transformation, and DWT features next to the extracted features to be combined for feature extraction. Then the optimum function is found with the Rider Optimization Algorithm (ROA) for all MLBP, GLCM and DWT. Eventually, the approach suggested is accepted. Deep Neural Network (DNN) performs the proposed authentication process. A DNN is a multilayered artificial neural network between the layers of input and output. The DNN finds the right mathematical manipulation to turn the input into the output, whether it is an acknowledged image or not. Suggested process quality is measured in terms of reliability recognition. In the MATLAB platform, the suggested approach is implemented.

Comparison of Input Image Dimensions for Ship Detection from KOMPSAT-5 SAR Image Using Deep Neural Network

2020 ◽  
Vol 102 (sp1) ◽  
Author(s):  
Wook Park ◽  
Won-Kyung Baek ◽  
Joong-Sun Won ◽  
Hyung-Sup Jung
Keyword(s):  
Neural Network ◽  
Deep Neural Network ◽  
Input Image ◽  
Sar Image ◽  
Ship Detection

scholarly journals A Multi-Column Deep Framework for Recognizing Artistic Media

Electronics ◽  
2019 ◽  
Vol 8 (11) ◽  
pp. 1277
Author(s):  
Yang ◽  
Min
Keyword(s):  
Neural Network ◽  
Computer Graphics ◽  
Deep Neural Network ◽  
Input Image ◽  
Local Scale ◽  
Gram Matrix ◽  
Final Decision ◽  
Texture Information ◽  
Artistic Medium ◽  
Image Datasets

We present a multi-column structured framework for recognizing artistic media from artwork images. We design the column of our framework using a deep neural network. Our key idea is to recognize the distinctive stroke texture of an artistic medium, which plays a key role in distinguishing artistic media. Since stroke texture is in a local scale, the whole image is not proper for recognizing the texture. Therefore, we devise two ideas for our framework: Sampling patches from an input image and employing a Gram matrix to extract the texture. The patches sampled from an input artwork image are processed in the columns of our framework to make local decisions on the patch, and the local decisions from the patches are merged to make a final decision for the input artwork image. Furthermore, we employ a Gram matrix, which is known to effectively capture texture information, to improve the accuracy of recognition. Our framework is trained and tested using two real artwork image datasets: WikiSet of traditional artwork images and YMSet of contemporary artwork images. Finally, we build SynthSet, which is a collection of synthesized artwork images from many computer graphics literature, and propose a guideline for evaluating the synthesized artwork images.

scholarly journals Automatic Academic Paper Rating Based on Convolutional Neural Network

2019 ◽  
pp. 569-574
Author(s):  
Mubeena A. K ◽  
Shahad P.
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Word Embedding ◽  
Second Phase ◽  
Personal Factors ◽  
Rating Score ◽  
Academic Paper ◽  
Proposed Model ◽  
Two Phases ◽  
Source Paper

As an ever increasing number of academic papers are being submitted to journals and conferences, assessing every one of these papers by experts is tedious and can cause imbalance because of the personal factors of the reviewers. In this system, in order to help professionals in assessing academic papers, here propose a task: Automatic Academic Paper Rating (AAPR), which automatically determine whether to accept academic papers. We build a convolutional neural network (CNN) model to achieve automatic academic paper rating task. It has two phases, first phase is identifying abstract part of source paper and generate rating score using CNN model and second phase is taking decision based on the score to accept or decline papers. This model takes word embedding of the abstracts as the input and learns useful features. The word embedding used for training the model is a semantically enriched set of Word2Vec word embedding. After the training phase, the proposed model will be able to generate the score of a new abstract. And find that the title and abstract parts have the most influence on whether the source paper quality when setting aside the other part of source papers. The proposed system outperforms the state-of-art technique.

scholarly journals Simultaneous Monocular Visual Odometry and Depth Reconstruction with Scale Recovery

2020 ◽  
Author(s):  
Guoliang Liu
Keyword(s):  
Neural Network ◽  
Deep Neural Network ◽  
Visual Odometry ◽  
Input Image ◽  
View Synthesis ◽  
Depth Image ◽  
The Core ◽  
Depth Prediction ◽  
Camera Pose ◽  
Depth Reconstruction

In this paper, we propose a deep neural networkthat can estimate camera poses and reconstruct thefull resolution depths of the environment simultaneously usingonly monocular consecutive images. In contrast to traditionalmonocular visual odometry methods, which cannot estimatescaled depths, we here demonstrate the recovery of the scaleinformation using a sparse depth image as a supervision signalin the training step. In addition, based on the scaled depth,the relative poses between consecutive images can be estimatedusing the proposed deep neural network. Another novelty liesin the deployment of view synthesis, which can synthesize anew image of the scene from a different view (camera pose)given an input image. The view synthesis is the core techniqueused for constructing a loss function for the proposed neuralnetwork, which requires the knowledge of the predicted depthsand relative poses, such that the proposed method couples thevisual odometry and depth prediction together. In this way,both the estimated poses and the predicted depths from theneural network are scaled using the sparse depth image as thesupervision signal during training. The experimental results onthe KITTI dataset show competitive performance of our methodto handle challenging environments.<br>

scholarly journals Machine Learning Can Predict Deaths in Patients with Diverticulitis During their Hospital Stay

2020 ◽  
Author(s):  
Fahad Shabbir Ahmed ◽  
Raza-Ul-Mustafa ◽  
Liaqat Ali ◽  
Imad-ud-Deen ◽  
Tahir Hameed ◽  
...  
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Risk Factors ◽  
At Risk ◽  
Logistic Regression ◽  
Deep Neural Network ◽  
Statistical Significance ◽  
Significant Risk ◽  
Premature Death ◽  
Second Phase

ABSTRACTIntroductionDiverticulitis is the inflammation and/or infection of small pouches known as diverticula that develop along the walls of the intestines. Patients with diverticulitis are at risk of mortality as high as 17% with abscess formation and 45% with secondary perforation, especially patients that get admitted to the inpatient services are at risk of complications including mortality. We developed a deep neural networks (DNN) based machine learning framework that could predict premature death in patients that are admitted with diverticulitis using electronic health records (EHR) to calculate the statistically significant risk factors first and then to apply deep neural network.MethodsOur proposed framework (Deep FLAIM) is a two-phase hybrid works framework. In the first phase, we used National In-patient Sample 2014 dataset to extract patients with diverticulitis patients with and without hemorrhage with the ICD-9 codes 562.11 and 562.13 respectively and analyzed these patients for different risk factors for statistical significance with univariate and multivariate analyses to generate hazard ratios, to rank the diverticulitis associated risk factors. In the second phase, we applied deep neural network model to predict death. Additionally, we have compared the performance of our proposed system by using the popular machine learning models such as DNN and Logistic Regression (LR).ResultsA total of 128,258 patients were used, we tested 64 different variables for using univariate and multivariate (age, gender and ethnicity) cox-regression for significance only 16 factors were statistically significant for both univariate and multivariate analysis. The mortality prediction for our DNN out-performed the conventional machine learning (logistic regression) in terms of AUC (0.977 vs 0.904), training accuracy (0.931 vs 0.900), testing accuracy (0.930 vs 0.910), sensitivity (90% vs 88%) and specificity (95% vs 93%).ConclusionOur Deep FLAIM Framework can predict mortality in patients admitted to the hospital with diverticulitis with high accuracy. The proposed framework can be expanded to predict premature death for other disease.

scholarly journals Research on the application of target detection based on deep learning technology in power grid operation inspection

2021 ◽  
Vol 236 ◽  
pp. 01035
Author(s):  
Peng Weifu ◽  
Du Shu ◽  
Chen Shaolei ◽  
Zhou Qing ◽  
Tang Na
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Power System ◽  
Target Detection ◽  
Early Warning ◽  
Deep Neural Network ◽  
Detection Method ◽  
Video Data ◽  
Detection Speed ◽  
External Damage

-External damage to power facilities caused by crane, excavator and other construction operations increases year by year, which will seriously threaten the safe operation of power system. It is an important measure to ensure the safe and reliable operation of power system to implement intelligent monitoring and early warning of power external breakdown through video and other non-contact observation means. The video data of power mainly comes from the fixed monitoring of helicopters, uavs and transformation poles and towers, which is characterized by large amount of data, complex scenes and serious environmental interference. The traditional target detection method usually selects the candidate area first, and then makes judgment based on the characteristics of human construction. The detection speed is slow and the accuracy is low, which makes it impossible to monitor the video data in real time, so as to make timely and accurate early warning and intervention fbr external damage. The target detection method based on deep learning optimizes or even eliminates the selection of candidate regions, which greatly speeds up the detection speed. By learning a lot of target samples through the deep neural network, the characteristics of high robustness are gradually fitted to make the target judgment more accurate. There are three key problems in introducing the target detection method based on deep learning into the power video detection: Firstly, the target detection method based on deep learning has a large amount of calculation and many parameters. In order to realize in-place operation on terminals with limited computing and storage capacity, it is necessary to find a practical method to simplify the network and reduce the amount of operational data in the detection process, which is the key to realize in-place operation and terminal operation of deep neural network. Secondly, for specific application scenarios, the effect of different target detection algorithms varies greatly, and there is a strong particularity of power video. Finding an effective target detection method is the key to improve the detection speed and accuracy. Finally, with the continuous development of deep learning, the structure of deep neural network changes with each passing day, and each has its own characteristics, which network structure is used as the feature extraction layer of target detection algorithm is the focus of research.

scholarly journals Deep neural network models reveal interplay of peripheral coding and stimulus statistics in pitch perception

2020 ◽  
Author(s):  
Mark R. Saddler ◽  
Ray Gonzalez ◽  
Josh H. McDermott
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Neural Coding ◽  
Deep Neural Network ◽  
Network Models ◽  
Pitch Perception ◽  
Natural Environments ◽  
Sensory Receptors ◽  
Natural Sounds ◽  
Neural Network Models

ABSTRACTComputations on receptor responses enable behavior in the environment. Behavior is plausibly shaped by both the sensory receptors and the environments for which organisms are optimized, but their roles are often opaque. One classic example is pitch perception, whose properties are commonly linked to peripheral neural coding limits rather than environmental acoustic constraints. We trained artificial neural networks to estimate fundamental frequency from simulated cochlear representations of natural sounds. The best-performing networks replicated many characteristics of human pitch judgments. To probe how our ears and environment shape these characteristics, we optimized networks given altered cochleae or sound statistics. Human-like behavior emerged only when cochleae had high temporal fidelity and when models were optimized for natural sounds. The results suggest pitch perception is critically shaped by the constraints of natural environments in addition to those of the cochlea, illustrating the use of contemporary neural networks to reveal underpinnings of behavior.

Sign in / Sign up

Export Citation Format

Share Document