scholarly journals MFAN: Multi-Level Features Attention Network for Fake Certificate Image Detection

Entropy ◽  
2022 ◽  
Vol 24 (1) ◽  
pp. 118
Author(s):  
Yu Sun ◽  
Rongrong Ni ◽  
Yao Zhao
Keyword(s):  
Contextual Information ◽  
Feature Maps ◽  
Attention Network ◽  
Final Layer ◽  
Spatial Pyramid Pooling ◽  
Novel Method ◽  
Feature Attention ◽  
Multi Level ◽  
The One ◽  
Small Targets

Up to now, most of the forensics methods have attached more attention to natural content images. To expand the application of image forensics technology, forgery detection for certificate images that can directly represent people’s rights and interests is investigated in this paper. Variable tampered region scales and diverse manipulation types are two typical characteristics in fake certificate images. To tackle this task, a novel method called Multi-level Feature Attention Network (MFAN) is proposed. MFAN is built following the encoder–decoder network structure. In order to extract features with rich scale information in the encoder, on the one hand, we employ Atrous Spatial Pyramid Pooling (ASPP) on the final layer of a pre-trained residual network to capture the contextual information at different scales; on the other hand, low-level features are concatenated to ensure the sensibility to small targets. Furthermore, the resulting multi-level features are recalibrated on channels for irrelevant information suppression and enhancing the tampered regions, guiding the MFAN to adapt to diverse manipulation traces. In the decoder module, the attentive feature maps are convoluted and unsampled to effectively generate the prediction mask. Experimental results indicate that the proposed method outperforms some state-of-the-art forensics methods.

scholarly journals MFFA-SARNET: Deep Transferred Multi-Level Feature Fusion Attention Network with Dual Optimized Loss for Small-Sample SAR ATR

2020 ◽  
Vol 12 (9) ◽  
pp. 1385
Author(s):  
Yikui Zhai ◽  
Wenbo Deng ◽  
Tian Lan ◽  
Bing Sun ◽  
Zilu Ying ◽  
...  
Keyword(s):  
Synthetic Aperture Radar ◽  
Feature Fusion ◽  
Small Sample ◽  
Classification Model ◽  
Synthetic Aperture ◽  
Sar Images ◽  
Attention Network ◽  
Feature Attention ◽  
Multi Level ◽  
Aperture Radar

Synthetic Aperture Radar (SAR) Automatic Target Recognition (ATR), most algorithms of which have employed and relied on sufficient training samples to receive a strong discriminative classification model, has remained a challenging task in recent years, among which the challenge of SAR data acquisition and further insight into the intuitive features of SAR images are the main concerns. In this paper, a deep transferred multi-level feature fusion attention network with dual optimized loss, called a multi-level feature attention Synthetic Aperture Radar network (MFFA-SARNET), is proposed to settle the problem of small samples in SAR ATR tasks. Firstly, a multi-level feature attention (MFFA) network is established to learn more discriminative features from SAR images with a fusion method, followed by alleviating the impact of background features on images with the following attention module that focuses more on the target features. Secondly, a novel dual optimized loss is incorporated to further optimize the classification network, which enhances the robust and discriminative learning power of features. Thirdly, transfer learning is utilized to validate the variances and small-sample classification tasks. Extensive experiments conducted on a public database with three different configurations consistently demonstrate the effectiveness of our proposed network, and the significant improvements yielded to surpass those of the state-of-the-art methods under small-sample conditions.

scholarly journals Multi-Resolution Image Segmentation Based on a Cascaded U-ADenseNet for the Liver and Tumors

2021 ◽  
Vol 11 (10) ◽  
pp. 1044
Author(s):  
Yan Zhu ◽  
Aihong Yu ◽  
Huan Rong ◽  
Dongqing Wang ◽  
Yuqing Song ◽  
...  
Keyword(s):  
Liver Tumors ◽  
Malignant Tumors ◽  
Contextual Information ◽  
Automatic Segmentation ◽  
Tumor Segmentation ◽  
Proposed Model ◽  
Wide Range ◽  
Spatial Pyramid Pooling ◽  
The One ◽  
Coarse To Fine

The liver is an irreplaceable organ in the human body, maintaining life activities and metabolism. Malignant tumors of the liver have a high mortality rate at present. Computer-aided segmentation of the liver and tumors has significant effects on clinical diagnosis and treatment. There are still many challenges in the segmentation of the liver and liver tumors simultaneously, such as, on the one hand, that convolutional kernels with fixed geometric structures do not match complex, irregularly shaped targets; on the other, pooling during convolution results in a loss of spatial contextual information of images. In this work, we designed a cascaded U-ADenseNet with coarse-to-fine processing for addressing the above issues of fully automatic segmentation. This work contributes multi-resolution input images and multi-layered channel attention combined with atrous spatial pyramid pooling densely connected in the fine segmentation. The proposed model was evaluated by a public dataset of the Liver Tumor Segmentation Challenge (LiTS). Our approach attained competitive liver and tumor segmentation scores that exceeded other methods across a wide range of metrics.

scholarly journals Design and Analysis of a Lightweight Context Fusion CNN Scheme for Crowd Counting

Sensors ◽  
2019 ◽  
Vol 19 (9) ◽  
pp. 2013 ◽  
Author(s):  
Yang Yu ◽  
Jifeng Huang ◽  
Wen Du ◽  
Naixue Xiong
Keyword(s):  
Resource Constraints ◽  
Contextual Information ◽  
Basic Feature ◽  
Traffic Monitoring ◽  
Feature Maps ◽  
Crowd Counting ◽  
Density Maps ◽  
Urban Security ◽  
Novel Method ◽  
Improving Accuracy

Crowd counting, which is widely used in disaster management, traffic monitoring, and other fields of urban security, is a challenging task that is attracting increasing interest from researchers. For better accuracy, most methods have attempted to handle the scale variation explicitly. which results in huge scale changes of the object size. However, earlier methods based on convolutional neural networks (CNN) have focused primarily on improving accuracy while ignoring the complexity of the model. This paper proposes a novel method based on a lightweight CNN-based network for estimating crowd counting and generating density maps under resource constraints. The network is composed of three components: a basic feature extractor (BFE), a stacked à trous convolution module (SACM), and a context fusion module (CFM). The BFE encodes basic feature information with reduced spatial resolution for further refining. Various pieces of contextual information are generated through a short pipeline in SACM. To generate a context fusion density map, CFM distills feature maps from the above components. The whole network is trained in an end-to-end fashion and uses a compression factor to restrict its size. Experiments on three highly-challenging datasets demonstrate that the proposed method delivers attractive performance.

scholarly journals Remote Sensing Image Super-Resolution Based on Dense Channel Attention Network

2021 ◽  
Vol 13 (15) ◽  
pp. 2966
Author(s):  
Yunchuan Ma ◽  
Pengyuan Lv ◽  
Hao Liu ◽  
Xuehong Sun ◽  
Yanfei Zhong
Keyword(s):  
Remote Sensing ◽  
High Frequency ◽  
Super Resolution ◽  
Remote Sensing Images ◽  
Feature Maps ◽  
Discriminative Ability ◽  
Attention Network ◽  
Feature Information ◽  
Image Super Resolution ◽  
Multi Level

In the recent years, convolutional neural networks (CNN)-based super resolution (SR) methods are widely used in the field of remote sensing. However, complicated remote sensing images contain abundant high-frequency details, which are difficult to capture and reconstruct effectively. To address this problem, we propose a dense channel attention network (DCAN) to reconstruct high-resolution (HR) remote sensing images. The proposed method learns multi-level feature information and pays more attention to the important and useful regions in order to better reconstruct the final image. Specifically, we construct a dense channel attention mechanism (DCAM), which densely uses the feature maps from the channel attention block via skip connection. This mechanism makes better use of multi-level feature maps which contain abundant high-frequency information. Further, we add a spatial attention block, which makes the network have more flexible discriminative ability. Experimental results demonstrate that the proposed DCAN method outperforms several state-of-the-art methods in both quantitative evaluation and visual quality.

A Novel Method for the Rapid Purification of Human and Rat Fibrin(OGEN) Degradation Products in High Yields

1979 ◽  
Author(s):  
W. Nieuwenhuizen ◽  
I. A. M. van Ruijven-Vermeer ◽  
F. Haverkate ◽  
G. Timan
Keyword(s):  
Degradation Products ◽  
Complete Separation ◽  
Purification Method ◽  
Amino Terminal ◽  
Novel Method ◽  
Rapid Purification ◽  
The One ◽  
High Yields ◽  
Size Heterogeneity ◽  
D Dimer

A novel method will be described for the preparation and purification of fibrin(ogen) degradation products in high yields. The high yields are due to two factors. on the one hand an improved preparation method in which the size heterogeneity of the degradation products D is strongly reduced by plasmin digestion at well-controlled calcium concentrations. At calcium concentrations of 2mM exclusively D fragments, M.W.= 93-000 (Dcate) were formed; in the presence of 1OmM EGTA only fragments M.W.= 80.000 (D EGTA) were formed as described. on the other hand a new purification method, which includes Sephadex G-200 filtration to purify the D:E complexes and separation of the D and E fragments by a 16 hrs. preparative isoelectric focussing. The latter step gives a complete separation of D (fragments) (pH = 6.5) and E fragments (at pH = 4.5) without any overlap, thus allowing a nearly 100% recovery in this step. The overall recoveries are around 75% of the theoretical values. These recoveries are superior to those of existing procedures. Moreover the conditions of this purification procedure are very mild and probably do not affect the native configuration of the products. Amino-terminal amino acids of human Dcate, D EGTA and D-dimer are identical i.e. val, asx and ser. in the ratgly, asx and ser were found. E 1% for rat Dcate=17-8 for rat D EGTA=16.2 and for rat D- dimer=l8.3. for the corresponding human fragments, these values were all 20.0 ± 0.2.

Near-Perfect Reconstruction Oversampled Nonuniform Cosine-Modulated Filter Banks Based on Frequency Warping and Subband Merging

2012 ◽  
Vol 58 (2) ◽  
pp. 177-192 ◽  
Author(s):  
Marek Parfieniuk ◽  
Alexander Petrovsky
Keyword(s):  
Group Delay ◽  
Filter Banks ◽  
Main Idea ◽  
Perceptual Processing ◽  
Perfect Reconstruction ◽  
Frequency Warping ◽  
Small Loss ◽  
Novel Method ◽  
The One ◽  
Near Perfect Reconstruction

Near-Perfect Reconstruction Oversampled Nonuniform Cosine-Modulated Filter Banks Based on Frequency Warping and Subband MergingA novel method for designing near-perfect reconstruction oversampled nonuniform cosine-modulated filter banks is proposed, which combines frequency warping and subband merging, and thus offers more flexibility than known techniques. On the one hand, desirable frequency partitionings can be better approximated. On the other hand, at the price of only a small loss in partitioning accuracy, both warping strength and number of channels before merging can be adjusted so as to minimize the computational complexity of a system. In particular, the coefficient of the function behind warping can be constrained to be a negative integer power of two, so that multiplications related to allpass filtering can be replaced with more efficient binary shifts. The main idea is accompanied by some contributions to the theory of warped filter banks. Namely, group delay equalization is thoroughly investigated, and it is shown how to avoid significant aliasing by channel oversampling. Our research revolves around filter banks for perceptual processing of sound, which are required to approximate the psychoacoustic scales well and need not guarantee perfect reconstruction.

Multi-level 3D Densenets for False-positive Reduction in Lung Nodule Detection Based on Chest Computed Tomography

2020 ◽  
Vol 16 (8) ◽  
pp. 1004-1021
Author(s):  
Xiaoqi Lu ◽  
Yu Gu ◽  
Lidong Yang ◽  
Baohua Zhang ◽  
Ying Zhao ◽  
...  
Keyword(s):  
False Positive ◽  
Contextual Information ◽  
Lung Nodule ◽  
Positive Sample ◽  
Lung Nodules ◽  
Lung Nodule Detection ◽  
Nodule Detection ◽  
Level Information ◽  
Multi Level ◽  
Cade System

Objective: False-positive nodule reduction is a crucial part of a computer-aided detection (CADe) system, which assists radiologists in accurate lung nodule detection. In this research, a novel scheme using multi-level 3D DenseNet framework is proposed to implement false-positive nodule reduction task. Methods: Multi-level 3D DenseNet models were extended to differentiate lung nodules from falsepositive nodules. First, different models were fed with 3D cubes with different sizes for encoding multi-level contextual information to meet the challenges of the large variations of lung nodules. In addition, image rotation and flipping were utilized to upsample positive samples which consisted of a positive sample set. Furthermore, the 3D DenseNets were designed to keep low-level information of nodules, as densely connected structures in DenseNet can reuse features of lung nodules and then boost feature propagation. Finally, the optimal weighted linear combination of all model scores obtained the best classification result in this research. Results: The proposed method was evaluated with LUNA16 dataset which contained 888 thin-slice CT scans. The performance was validated via 10-fold cross-validation. Both the Free-response Receiver Operating Characteristic (FROC) curve and the Competition Performance Metric (CPM) score show that the proposed scheme can achieve a satisfactory detection performance in the falsepositive reduction track of the LUNA16 challenge. Conclusion: The result shows that the proposed scheme can be significant for false-positive nodule reduction task.

The Metaphysics of Representation

2020 ◽  
Author(s):  
J. Robert G. Williams
Keyword(s):  
Representational Content ◽  
Physical World ◽  
Degrees Of Belief ◽  
Public Language ◽  
Language And Thought ◽  
Final Layer ◽  
The World ◽  
Interactive Relationship ◽  
Compositional Interpretation ◽  
The One

What is representation? How do the more primitive aspects of our world come together to generate it? How do different kinds of representation relate to one another? This book identifies the metaphysical foundations for representational facts. The story told is in three parts. The most primitive layer of representation is the ‘aboutness’ of sensation/perception and intention/action, which are the two most basic modes in which an individual and the world interact. It is argued that we can understand how this kind of representation can exist in a fundamentally physical world so long as we have an independent, illuminating grip on functions and causation. The second layer of representation is the ‘aboutness’ of (degrees of) belief and desire, whose representational content goes far beyond the immediate perceptable and manipulable environment. It is argued that the correct belief/desire interpretation of an agent is the one which makes their action-guiding states, given their perceptual evidence, most rational. The final layer of representation is the ‘aboutness’ of words and sentences, human artefacts with representational content. It is argued that one can give an illuminating account of the conditions under which a compositional interpretation of a public language like English is correct by appeal to patterns emerging from the attitudes conventionally expressed by sentences. The three-layer metaphysics of representation resolves long-standing underdetermination puzzles, predicts and explains patterns in the way that concepts denote, and articulates a delicate interactive relationship between the foundations of language and thought.

scholarly journals A novel approach to the computation of one-loop three- and four-point functions: I. The real mass case

2019 ◽  
Vol 2019 (11) ◽  
Author(s):  
J Ph Guillet ◽  
E Pilon ◽  
Y Shimizu ◽  
M S Zidi
Keyword(s):  
Building Blocks ◽  
The Real ◽  
Alternative Method ◽  
Novel Approach ◽  
Reduction Methods ◽  
Real Mass ◽  
Novel Method ◽  
Algebraic Reduction ◽  
The One ◽  
Mass Case

Abstract This article is the first of a series of three presenting an alternative method of computing the one-loop scalar integrals. This novel method enjoys a couple of interesting features as compared with the method closely following ’t Hooft and Veltman adopted previously. It directly proceeds in terms of the quantities driving algebraic reduction methods. It applies to the three-point functions and, in a similar way, to the four-point functions. It also extends to complex masses without much complication. Lastly, it extends to kinematics more general than that of the physical, e.g., collider processes relevant at one loop. This last feature may be useful when considering the application of this method beyond one loop using generalized one-loop integrals as building blocks.

scholarly journals Attention Modulated Multiple Object Tracking with Motion Enhancement and Dual Correlation

Symmetry ◽  
2021 ◽  
Vol 13 (2) ◽  
pp. 266 ◽  
Author(s):  
Yifeng Wang ◽  
Zhijiang Zhang ◽  
Ning Zhang ◽  
Dan Zeng
Keyword(s):  
Object Tracking ◽  
Multiple Object Tracking ◽  
Tracking Accuracy ◽  
Feature Maps ◽  
Backbone Networks ◽  
Multiple Object ◽  
The Arts ◽  
Training Stage ◽  
The One ◽  
Baseline State

The one-shot multiple object tracking (MOT) framework has drawn more and more attention in the MOT research community due to its advantage in inference speed. However, the tracking accuracy of current one-shot approaches could lead to an inferior performance compared with their two-stage counterparts. The reasons are two-fold: one is that motion information is often neglected due to the single-image input. The other is that detection and re-identification (ReID) are two different tasks with different focuses. Joining detection and re-identification at the training stage could lead to a suboptimal performance. To alleviate the above limitations, we propose a one-shot network named Motion and Correlation-Multiple Object Tracking (MAC-MOT). MAC-MOT introduces a motion enhance attention module (MEA) and a dual correlation attention module (DCA). MEA performs differences on adjacent feature maps which enhances the motion-related features while suppressing irrelevant information. The DCA module focuses on decoupling the detection task and re-identification task to strike a balance and reduce the competition between these two tasks. Moreover, symmetry is a core design idea in our proposed framework which is reflected in Siamese-based deep learning backbone networks, the input of dual stream images, as well as a dual correlation attention module. Our proposed approach is evaluated on the popular multiple object tracking benchmarks MOT16 and MOT17. We demonstrate that the proposed MAC-MOT can achieve a better performance than the baseline state of the arts (SOTAs).

Sign in / Sign up

Export Citation Format

Share Document