scholarly journals Accelerated Correction of Reflection Artifacts by Deep Neural Networks in Photo-Acoustic Tomography

2019 ◽  
Vol 9 (13) ◽  
pp. 2615 ◽  
Author(s):  
Hongming Shan ◽  
Ge Wang ◽  
Yang Yang
Keyword(s):  
Diagnostic Performance ◽  
Deep Neural Networks ◽  
State Of The Art ◽  
Signal To Noise Ratio ◽  
Iterative Algorithms ◽  
Structural Similarity ◽  
Acoustic Tomography ◽  
Acoustic Reflection ◽  
Non Invasive ◽  
Imaging Performance

Photo-Acoustic Tomography (PAT) is an emerging non-invasive hybrid modality driven by a constant yearning for superior imaging performance. The image quality, however, hinges on the acoustic reflection, which may compromise the diagnostic performance. To address this challenge, we propose to incorporate a deep neural network into conventional iterative algorithms to accelerate and improve the correction of reflection artifacts. Based on the simulated PAT dataset from computed tomography (CT) scans, this network-accelerated reconstruction approach is shown to outperform two state-of-the-art iterative algorithms in terms of the peak signal-to-noise ratio (PSNR) and the structural similarity (SSIM) in the presence of noise. The proposed network also demonstrates considerably higher computational efficiency than conventional iterative algorithms, which are time-consuming and cumbersome.

scholarly journals Blind Color Image Watermarking Using Fan Beam Transform and QR Decomposition

Symmetry ◽  
2020 ◽  
Vol 12 (3) ◽  
pp. 486
Author(s):  
Pranab Kumar Dhar ◽  
Pulak Hazra ◽  
Tetsuya Shimamura
Keyword(s):  
Color Image ◽  
State Of The Art ◽  
Signal To Noise Ratio ◽  
Image Watermarking ◽  
Triangular Matrix ◽  
Structural Similarity ◽  
Qr Decomposition ◽  
Original Image ◽  
Color Image Watermarking ◽  
Multimedia Contents

Digital watermarking has been utilized effectively for copyright protection of multimedia contents. This paper suggests a blind symmetric watermarking algorithm using fan beam transform (FBT) and QR decomposition (QRD) for color images. At first, the original image is transferred from RGB to L*a*b* color model and FBT is applied to b* component. Then the b*component of the original image is split into m × m non-overlapping blocks and QRD is conducted to each block. Watermark data is placed into the selected coefficient of the upper triangular matrix using a new embedding function. Simulation results suggest that the presented algorithm is extremely robust against numerous attacks, and also yields watermarked images with high quality. Furthermore, it represents more excellent performance compared with the recent state-of-the-art algorithms for robustness and imperceptibility. The normalized correlation (NC) of the proposed algorithm varies from 0.8252 to 1, the peak signal-to-noise ratio (PSNR) varies from 54.1854 to 54.1892, and structural similarity (SSIM) varies from 0.9285 to 0.9696, respectively. In contrast, the NC of the recent state-of-the-art algorithms varies from 0.5193 to 1, PSNR varies from 38.5471 to 52.64, and SSIM varies from 0.9311 to 0.9663, respectively.

scholarly journals Spoofing Speaker Verification System by Adversarial Examples Leveraging the Generalized Speaker Difference

2021 ◽  
Vol 2021 ◽  
pp. 1-10
Author(s):  
Hongwei Luo ◽  
Yijie Shen ◽  
Feng Lin ◽  
Guoai Xu
Keyword(s):  
Neural Networks ◽  
Loss Function ◽  
Deep Neural Networks ◽  
State Of The Art ◽  
Speaker Verification ◽  
Signal To Noise Ratio ◽  
The State ◽  
Verification System ◽  
Adversarial Examples ◽  
Human Hearing

Speaker verification system has gained great popularity in recent years, especially with the development of deep neural networks and Internet of Things. However, the security of speaker verification system based on deep neural networks has not been well investigated. In this paper, we propose an attack to spoof the state-of-the-art speaker verification system based on generalized end-to-end (GE2E) loss function for misclassifying illegal users into the authentic user. Specifically, we design a novel loss function to deploy a generator for generating effective adversarial examples with slight perturbation and then spoof the system with these adversarial examples to achieve our goals. The success rate of our attack can reach 82% when cosine similarity is adopted to deploy the deep-learning-based speaker verification system. Beyond that, our experiments also reported the signal-to-noise ratio at 76 dB, which proves that our attack has higher imperceptibility than previous works. In summary, the results show that our attack not only can spoof the state-of-the-art neural-network-based speaker verification system but also more importantly has the ability to hide from human hearing or machine discrimination.

scholarly journals Raindrop-Aware GAN: Unsupervised Learning for Raindrop-Contaminated Coastal Video Enhancement

2020 ◽  
Vol 12 (20) ◽  
pp. 3461
Author(s):  
Jinah Kim ◽  
Dong Huh ◽  
Taekyung Kim ◽  
Jaeil Kim ◽  
Jeseon Yoo ◽  
...  
Keyword(s):  
State Of The Art ◽  
Signal To Noise Ratio ◽  
Structural Similarity ◽  
Visual Assessment ◽  
Background Information ◽  
Video Enhancement ◽  
Adversarial Learning ◽  
Video Images ◽  
Degraded Image

We propose an unsupervised network with adversarial learning, the Raindrop-aware GAN, which enhances the quality of coastal video images contaminated by raindrops. Raindrop removal from coastal videos faces two main difficulties: converting the degraded image into a clean one by visually removing the raindrops, and restoring the background coastal wave information in the raindrop regions. The components of the proposed network—a generator and a discriminator for adversarial learning—are trained on unpaired images degraded by raindrops and clean images free from raindrops. By creating raindrop masks and background-restored images, the generator restores the background information in the raindrop regions alone, preserving the input as much as possible. The proposed network was trained and tested on an open-access dataset and directly collected dataset from the coastal area. It was then evaluated by three metrics: the peak signal-to-noise ratio, structural similarity, and a naturalness-quality evaluator. The indices of metrics are 8.2% (+2.012), 0.2% (+0.002), and 1.6% (−0.196) better than the state-of-the-art method, respectively. In the visual assessment of the enhanced video image quality, our method better restored the image patterns of steep wave crests and breaking than the other methods. In both quantitative and qualitative experiments, the proposed method more effectively removed the raindrops in coastal video and recovered the damaged background wave information than state-of-the-art methods.

scholarly journals Learning a Convolutional Autoencoder for Nighttime Image Dehazing

Information ◽  
2020 ◽  
Vol 11 (9) ◽  
pp. 424
Author(s):  
Mengyao Feng ◽  
Teng Yu ◽  
Mingtao Jing ◽  
Guowei Yang
Keyword(s):  
State Of The Art ◽  
Signal To Noise Ratio ◽  
Local Maximum ◽  
Structural Similarity ◽  
Light Sources ◽  
Ambient Illumination ◽  
Image Dehazing ◽  
Edge Preserving ◽  
Convolutional Autoencoder ◽  
Maximum Method

Currently, haze removal of images captured at night for foggy scenes rely on the traditional, prior-based methods, but these methods are frequently ineffective at dealing with night hazy images. In addition, the light sources at night are complicated and there is a problem of inconsistent brightness. This makes the estimation of the transmission map complicated in the night scene. Based on the above analysis, we propose an autoencoder method to solve the problem of overestimation or underestimation of transmission captured by the traditional, prior-based methods. For nighttime hazy images, we first remove the color effect of the haze image with an edge-preserving maximum reflectance prior (MRP) method. Then, the hazy image without color influence is input into the self-encoder network with skip connections to obtain the transmission map. Moreover, instead of using the local maximum method, we estimate the ambient illumination through a guiding image filtering. In order to highlight the effectiveness of our experiments, a large number of comparison experiments were conducted between our method and the state-of-the-art methods. The results show that our method can effectively suppress the halo effect and reduce the effectiveness of glow. In the experimental part, we calculate that the average Peak Signal to Noise Ratio (PSNR) is 21.0968 and the average Structural Similarity (SSIM) is 0.6802.

Making Synchrotron Tomography a Routine Tool for 3D Integration Failure Analysis through a Limited Number of Projections, an Adapted Sample Preparation Scheme, and a Fully-Automated Post-Processing

2017 ◽  
Author(s):  
A. Fraczkiewicz ◽  
S. Moreau ◽  
T. Mourier ◽  
P. Bleuet ◽  
P.-O. Autran ◽  
...  
Keyword(s):  
Sample Preparation ◽  
Signal To Noise Ratio ◽  
Iterative Algorithms ◽  
Structural Similarity ◽  
High Signal ◽  
Similarity Function ◽  
Post Processing ◽  
3D Integration ◽  
Whole Process ◽  
Reconstructed Volume

Abstract 3D integration takes more and more importance in the microelectronics industry. This paper focuses on two types or objects, which are copper pillars (25 micrometer of diameter) and hybrid bonding samples. It aims at a statistical morphology observation of hybrid bonding structures, which underwent an electromigration test at 350 deg C and 20 mA. The goal of the study is two-fold. It is both to limit the overall time needed to perform a whole process flow, from sample preparation to reconstructed volume, and to limit the time of human intervention. To achieve this goal, three strategies are presented: improving the sample preparation scheme, reducing the number of projections with iterative algorithms and the Structural SIMilarity function, and automating the post-processing. The post-processing of the data is fully automated and directly renders the reconstructed volume. The high signal to noise ratio allows for further segmentation and analysis.

scholarly journals A Deep Classifier for Upper-Limbs Motor Anticipation Tasks in an Online BCI Setting

2021 ◽  
Vol 8 (2) ◽  
pp. 21
Author(s):  
Andrea Valenti ◽  
Michele Barsotti ◽  
Davide Bacciu ◽  
Luca Ascari
Keyword(s):  
Real Time ◽  
Deep Neural Networks ◽  
State Of The Art ◽  
Brain Activity ◽  
Brain Computer Interface ◽  
Computer Interface ◽  
Non Invasive ◽  
The Brain ◽  
Movement Intention

Decoding motor intentions from non-invasive brain activity monitoring is one of the most challenging aspects in the Brain Computer Interface (BCI) field. This is especially true in online settings, where classification must be performed in real-time, contextually with the user’s movements. In this work, we use a topology-preserving input representation, which is fed to a novel combination of 3D-convolutional and recurrent deep neural networks, capable of performing multi-class continual classification of subjects’ movement intentions. Our model is able to achieve a higher accuracy than a related state-of-the-art model from literature, despite being trained in a much more restrictive setting and using only a simple form of input signal preprocessing. The results suggest that deep learning models are well suited for deployment in challenging real-time BCI applications such as movement intention recognition.

ANALISIS SENSITIVITAS VIDEO MPEG-4 BERDASARKAN STRUKTUR FRAME PADA TRANSMISI DVB-T

2020 ◽  
Vol 25 (2) ◽  
pp. 86-97
Author(s):  
Sandy Suryo Prayogo ◽  
Tubagus Maulana Kusuma
Keyword(s):  
Deep Learning ◽  
Bit Error Rate ◽  
Error Rate ◽  
Signal To Noise Ratio ◽  
Similarity Index ◽  
Structural Similarity ◽  
Signal To Noise ◽  
Structural Similarity Index ◽  
Noise Ratio

DVB merupakan standar transmisi televisi digital yang paling banyak digunakan saat ini. Unsur terpenting dari suatu proses transmisi adalah kualitas gambar dari video yang diterima setelah melalui proses transimisi tersebut. Banyak faktor yang dapat mempengaruhi kualitas dari suatu gambar, salah satunya adalah struktur frame dari video. Pada tulisan ini dilakukan pengujian sensitifitas video MPEG-4 berdasarkan struktur frame pada transmisi DVB-T. Pengujian dilakukan menggunakan simulasi matlab dan simulink. Digunakan juga ffmpeg untuk menyediakan format dan pengaturan video akan disimulasikan. Variabel yang diubah dari video adalah bitrate dan juga group-of-pictures (GOP), sedangkan variabel yang diubah dari transmisi DVB-T adalah signal-to-noise-ratio (SNR) pada kanal AWGN di antara pengirim (Tx) dan penerima (Rx). Hasil yang diperoleh dari percobaan berupa kualitas rata-rata gambar pada video yang diukur menggunakan metode pengukuran structural-similarity-index (SSIM). Dilakukan juga pengukuran terhadap jumlah bit-error-rate BER pada bitstream DVB-T. Percobaan yang dilakukan dapat menunjukkan seberapa besar sensitifitas bitrate dan GOP dari video pada transmisi DVB-T dengan kesimpulan semakin besar bitrate maka akan semakin buruk nilai kualitas gambarnya, dan semakin kecil nilai GOP maka akan semakin baik nilai kualitasnya. Penilitian diharapkan dapat dikembangkan menggunakan deep learning untuk memperoleh frame struktur yang tepat di kondisi-kondisi tertentu dalam proses transmisi televisi digital.

scholarly journals Non-invasive three-dimensional thickness analysis of oral epithelium based on optical coherence tomography—development and diagnostic performance

Heliyon ◽  
2021 ◽  
Vol 7 (4) ◽  
pp. e06645
Author(s):  
Charlotte Theresa Trebing ◽  
Sinan Sen ◽  
Stefan Rues ◽  
Christopher Herpel ◽  
Maria Schöllhorn ◽  
...  
Keyword(s):  
Optical Coherence Tomography ◽  
Diagnostic Performance ◽  
Three Dimensional ◽  
Oral Epithelium ◽  
Optical Coherence ◽  
Non Invasive

scholarly journals A72 NON-INVASIVE PREDICTION OF ESOPHAGEAL VARICES BY TRANSIENT ELASTOGRAPHY AND PLATELET COUNT IN PATIENTS WITH HEPATITIS B AND ADVANCED CHRONIC LIVER DISEASE: VALIDATION OF BAVENO VI AND EXPANDED BAVENO VI CRITERIA

2021 ◽  
Vol 4 (Supplement_1) ◽  
pp. 37-38
Author(s):  
A Zoughlami ◽  
J Serero ◽  
G Sebastiani ◽  
M Deschenes ◽  
P Wong ◽  
...  
Keyword(s):  
Platelet Count ◽  
Liver Disease ◽  
Hepatitis B ◽  
Chronic Liver Disease ◽  
Esophageal Varices ◽  
Diagnostic Performance ◽  
Transient Elastography ◽  
Chronic Liver ◽  
Non Invasive ◽  
Chronic Hepatitis B Virus

Abstract Background Patients with compensated advanced chronic liver disease (cACLD) are at higher risk of developing complications from portal hypertension, including esophageal varices (EV). Baveno VI and expanded Baveno VI criteria, based on liver stiffness measurement (LSM) by transient elastography combined with platelet count, have been proposed to avoid unnecessary esophagogastroduodenoscopy (EGD) screening for large esophageal varices needing treatment (EVNT). This approach has not been validated in patients with chronic hepatitis B virus (HBV) infection, who have etiology-specific cut-off of LSM for liver fibrosis. Aims We aimed to validate the Baveno VI and expanded Baveno VI criteria for EVNT in HBV patients with cACLD. Methods We performed a retrospective analysis of HBV patients who underwent LSM in 2014–2020. Inclusion criteria were: a) diagnosis of cACLD, defined as LSM >9 kPa; b) availability of EGD and platelets within 1 year of LSM. Baveno VI (LSM <20 kPa and platelets >150,000) and expanded Baveno VI criteria (LSM <25 kPa and platelets >110,000) were tested for EGD sparing. Diagnostic performance of these criteria against gold standard (EGD) was computed and compared to patients with hepatitis C virus (HCV) infection and nonalcoholic steatohepatitis (NASH) etiologies, where these criteria have been widely validated. In these patients, the threshold for cACLD definition was >10 kPa. Results A total of 287 patients (mean age 56, 95% Child A) were included, comprising of 43 HBV (58% on antiviral therapy), 134 HCV and 110 NASH patients. The prevalence of any grade EV and EVNT was 25% and 8% in the whole cohort, with 19% and 5% in HBV patients, respectively. Table 1 reports diagnostic performance, spared EGD and missed EVNT according to non-invasive criteria and cACLD etiology. Both Baveno VI and expanded Baveno VI criteria performed well in patients with HBV-related cACLD. There was no significant difference on diagnostic performance of these non-invasive criteria across the cACLD etiologies. Conclusions These results support use of non-invasive criteria based on LSM and platelets to spare unnecessary EGD in patients with HBV and cACLD. Baveno VI and expanded Baveno VI criteria can improve resource utilization and avoid invasive testing in context of screening EGD for patients with HBV-related cACLD. Funding Agencies None

Sign in / Sign up

Export Citation Format

Share Document