Search

An underwater image often suffers from quality degradation issues, such as color deviations, low contrast, and blurred details, due to the absorption and scattering of light. In this article, we propose to address the aforementioned degradation issues via attenuated color channel correction and detail preserved contrast enhancement. Concretely, we first propose an underwater image color correction method. Considering the differences between superior and inferior color channels of an underwater image, the inferior color channels are compensated via especially designed attenuation matrices. We then employ a dual-histogram-based iterative threshold method and a limited histogram method with Rayleigh distribution to improve the global and local contrast of the color-corrected image, thus achieving a global contrast-enhanced version and a local contrast-enhanced version, respectively. To integrate the complementary merits between the global contrast-enhanced version and the local contrast-enhanced version, we adopt a multiscale fusion strategy to fuse them. Finally, we propose a multiscale unsharp masking strategy to further sharpen the fused image for better visual quality. Extensive experiments on four underwater image enhancement benchmark data sets demonstrate that our method effectively enhances underwater images qualitatively and quantitatively. Besides, our method also generalizes well to the enhancement of low-light images and hazy images. [ABSTRACT FROM AUTHOR]

Image enhancement is a technique to enhance the illumination of dark images while keeping the reality and the naturalness of the enhanced images at the same time. For color images, most methods tackle different color channels in a separate way, which overlooks the connection between the color channels. Therefore, in this letter, we consider a quaternion-based model to reserve the color connectivity, which integrates the color information of a pixel by a quaternion number. Moreover, we propose a regularizer based on the gamma-correction function and incorporate it into a screened Poisson equation for the image enhancement task. The uniqueness and existence of the solution of the proposed model are analyzed. The numerical results also prove the superiority of our scheme for color image enhancement. [ABSTRACT FROM AUTHOR]

A problem of image denoising, when images are corrupted by a non-stationary noise, is considered in this paper. Since, in practice, no a priori information on noise is available, noise statistics should be pre-estimated prior to image denoising. In this paper, deep convolutional neural network (CNN) based method for estimation of a map of local, patch-wise, standard deviations of noise (so-called sigma-map) is proposed. It achieves the state-of-the-art performance in accuracy of estimation of sigma-map for the case of non-stationary noise, as well as estimation of a noise variance for the case of an additive white Gaussian noise. Extensive experiments on image denoising using estimated sigma-maps demonstrate that our method outperforms recent CNN-based blind image denoising methods by up to 6 dB in PSNR, as well as other state-of-the-art methods based on sigma-map estimation by up to 0.5 dB, providing, at the same time, better usage flexibility. A comparison with the ideal case, when denoising is applied using ground-truth sigma-map, shows that a difference of corresponding PSNR values for the most of noise levels is within 0.1-0.2 dB, and does not exceed 0.6 dB. [ABSTRACT FROM AUTHOR]

Recommendation accuracy is a fundamental problem in the quality of the recommendation system. In this article, we propose an efficient deep matrix factorization (EDMF) with review feature learning for the industrial recommender system. Two characteristics in user’s review are revealed. First, interactivity between the user and the item, which can also be considered as the former’s scoring behavior on the latter, is exploited in a review. Second, the review is only a partial description of the user’s preferences for the item, which is revealed as the sparsity property. Specifically, in the first characteristic, EDMF extracts the interactive features of onefold review by convolutional neural networks with word-attention mechanism. Subsequently, ${L}_{0}$ norm is leveraged to constrain the review considering that the review information is a sparse feature, which is the second characteristic. Furthermore, the loss function is constructed by maximum a posteriori estimation theory, where the interactivity and sparsity property are converted as two prior probability functions. Finally, the alternative minimization algorithm is introduced to optimize the loss functions. Experimental results on several datasets demonstrate that the proposed methods, which show good industrial conversion application prospects, outperform the state-of-the-art methods in terms of effectiveness and efficiency. [ABSTRACT FROM AUTHOR]

Applying image-based processing methods to original videos on a framewise level breaks the temporal consistency between consecutive frames. Traditional video temporal consistency methods reconstruct an original frame containing flickers from corresponding nonflickering frames, but the inaccurate correspondence realized by optical flow restricts their practical use. In this article, we propose a temporally broad learning system (TBLS), an approach that enforces temporal consistency between frames. We establish the TBLS as a flat network comprising the input data, consisting of an original frame in an original video, a corresponding frame in the temporally inconsistent video on which the image-based technique was applied, and an output frame of the last original frame, as mapped features in feature nodes. Then, we refine extracted features by enhancing the mapped features as enhancement nodes with randomly generated weights. We then connect all extracted features to the output layer with a target weight vector. With the target weight vector, we can minimize the temporal information loss between consecutive frames and the video fidelity loss in the output videos. Finally, we remove the temporal inconsistency in the processed video and output a temporally consistent video. Besides, we propose an alternative incremental learning algorithm based on the increment of the mapped feature nodes, enhancement nodes, or input data to improve learning accuracy by a broad expansion. We demonstrate the superiority of our proposed TBLS by conducting extensive experiments. [ABSTRACT FROM AUTHOR]

The retinal vasculature provides important clues in the diagnosis and monitoring of systemic diseases including hypertension and diabetes. The microvascular system is of primary involvement in such conditions, and the retina is the only anatomical site where the microvasculature can be directly observed. The objective assessment of retinal vessels has long been considered a surrogate biomarker for systemic vascular diseases, and with recent advancements in retinal imaging and computer vision technologies, this topic has become the subject of renewed attention. In this paper, we present a novel dataset, dubbed RAVIR, for the semantic segmentation of Retinal Arteries and Veins in Infrared Reflectance (IR) imaging. It enables the creation of deep learning-based models that distinguish extracted vessel type without extensive post-processing. We propose a novel deep learning-based methodology, denoted as SegRAVIR, for the semantic segmentation of retinal arteries and veins and the quantitative measurement of the widths of segmented vessels. Our extensive experiments validate the effectiveness of SegRAVIR and demonstrate its superior performance in comparison to state-of-the-art models. Additionally, we propose a knowledge distillation framework for the domain adaptation of RAVIR pretrained networks on color images. We demonstrate that our pretraining procedure yields new state-of-the-art benchmarks on the DRIVE, STARE, and CHASE_DB1 datasets. Dataset link: https://ravirdataset.github.io/data. [ABSTRACT FROM AUTHOR]

Recent digital pathology workflows mainly focus on mono-modality histopathology image analysis. However, they ignore the complementarity between Haematoxylin & Eosin (H&E) and Immunohistochemically (IHC) stained images, which can provide comprehensive gold standard for cancer diagnosis. To resolve this issue, we propose a cross-boosted multi-target domain adaptation pipeline for multi-modality histopathology images, which contains Cross-frequency Style-auxiliary Translation Network (CSTN) and Dual Cross-boosted Segmentation Network (DCSN). Firstly, CSTN achieves the one-to-many translation from fluorescence microscopy images to H&E and IHC images for providing source domain training data. To generate images with realistic color and texture, Cross-frequency Feature Transfer Module (CFTM) is developed to pertinently restructure and normalize high-frequency content and low-frequency style features from different domains. Then, DCSN fulfills multi-target domain adaptive segmentation, where a dual-branch encoder is introduced, and Bidirectional Cross-domain Boosting Module (BCBM) is designed to implement cross-modality information complementation through bidirectional inter-domain collaboration. Finally, we establish Multi-modality Thymus Histopathology (MThH) dataset, which is the largest publicly available H&E and IHC image benchmark. Experiments on MThH dataset and several public datasets show that the proposed pipeline outperforms state-of-the-art methods on both histopathology image translation and segmentation. [ABSTRACT FROM AUTHOR]

This work was conducted to study the Chinese painting color teaching analysis of intelligent image color processing technology under the network environment. First, the paper preprocesses the obtained color mural images, realizes the automatic recognition and marking of the images with different defect degrees and color fading, and uses denoising and texture background elimination to remove unnecessary background information. Then, according to the characteristic that the repair order of boundary points in the Criminisi algorithm is determined by the size of priority weight, the data items and confidence items are added. Finally, the design uses image processing technology and the loss formula to identify the connecting edge of the color area to be taught, establish the color extraction area, calculate the bit weight of the best color, find out the color extraction position, and synthesize different colors according to the original painting color superposition method. The partial differential equation is used to set the teaching code of color teaching system to realize the teaching of Chinese painting color. The experimental results show that compared with the original teaching system, the designed color teaching system has a stronger ability to recognize the edge of Chinese painting color teaching, and the quality of Chinese painting after teaching is higher. It can be seen that the color teaching system can be applied to the color teaching of Chinese painting. [ABSTRACT FROM AUTHOR]

It is common to have to process signals or images whose values are cyclic and can be represented as points on the complex circle, like wrapped phases, angles, orientations, or color hues. We consider a Tikhonov-type regularization model to smoothen or interpolate circle-valued signals defined on arbitrary graphs. We propose a convex relaxation of this nonconvex problem as a semidefinite program, and an efficient algorithm to solve it. [ABSTRACT FROM AUTHOR]

In recent years, generative adversarial networks (GANs) have been widely used to generate realistic fake face images, which can easily deceive human beings. To detect these images, some methods have been proposed. However, their detection performance will be degraded greatly when the testing samples are post-processed. In this paper, some experimental studies on detecting post-processed GAN-generated face images find that (a) both the luminance component and chrominance components play an important role, and (b) the RGB and YCbCr color spaces achieve better performance than the HSV and Lab color spaces. Therefore, to enhance the robustness, both the luminance component and chrominance components of dual-color spaces (RGB and YCbCr) are considered to utilize color information effectively. In addition, the convolutional block attention module and multilayer feature aggregation module are introduced into the Xception model to enhance its feature representation power and aggregate multilayer features, respectively. Finally, a robust dual-stream network is designed by integrating dual-color spaces RGB and YCbCr and using an improved Xception model. Experimental results demonstrate that our method outperforms some existing methods, especially in its robustness against different types of post-processing operations, such as JPEG compression, Gaussian blurring, gamma correction, and median filtering. [ABSTRACT FROM AUTHOR]

Single image haze removal, which is to recover the clear version of a hazy image, is a challenging task in computer vision. In this paper, an additive haze model is proposed to approximate the hazy image formation process. In contrast with the traditional optical model, it regards the haze as an additive layer to a clean image. The model thus avoids estimating the medium transmission rate and the global atmospherical light. In addition, based on a critical observation that haze changes gradually and smoothly across the image, a haze smoothness prior is proposed to constrain this model. This prior assumes that the haze layer is much smoother than the clear image. Benefiting from this prior, we can directly separate the clean image from a single hazy image. Experimental results and comparisons with synthetic images and real-world images demonstrate that the proposed method outperforms state-of-the-art single image haze removal algorithms. [ABSTRACT FROM AUTHOR]

Underwater captured images are usually degraded by low contrast, hazy, and blurry due to absorbing and scattering, which limits their analyses and applications. To address these problems, a red channel prior guided variational framework is proposed based on the complete underwater image formation model (UIFM). Unlike most of the existing methods that only consider the direct transmission and backscattering components, we additionally include forward scattering component into the UIFM. In the proposed variational framework, we successfully incorporate the normalized total variation item and sparse prior knowledge of blur kernel together. In addition, we perform the estimation of blur kernel by varying image resolution in a coarse-to-fine manner to avoid local minima. Moreover, for solving the generated non-smooth optimization problem, we employ the alternating direction method of multipliers (ADMM) to accelerate the whole progress. Experimental results demonstrate that the proposed method has a good performance on dehazing and deblurring. Extensive qualitative and quantitative comparisons further validate its superiority against the other state-of-the-art algorithms. The code is available online at: https://github.com/Hou-Guojia/UNTV [ABSTRACT FROM AUTHOR]

In image processing, smoke may degrade visibility and deteriorate the performance of high-level vision applications. Therefore, single image smoke removal is crucial for computer vision. Currently, existing smoke removal algorithms mainly leverage handcrafted priors. Moreover, these methods usually apply haze removal methods to perform smoke removal due to the similarity between smoke and haze. However, these methods cannot sufficiently address the degradation of thick smoke and may suffer from residual smoke and color distortion problems due to the non-global and non-homogeneous distribution of smoke. In this paper, to solve the aforementioned problems, an end-to-end deep neural network called DesmokeNet is proposed. We construct a two-stage recovered pipeline to remove the smoke in different thicknesses. The light and thick smoke is first removed locally by the smoke removal network (SRN). The missing pixels in the thick smoke are then recovered by the pixel compensation network (PCN). Moreover, we proposed the thickness-aware pixel loss and the dark channel loss to suppress the residual smoke. To further increase the discriminative ability of the DesmokeNet, we proposed self-attentive feature consensus loss and multi-level contrastive regularization loss to improve the performance of smoke removal. Finally, to train the proposed method, we construct the first large-scale dataset containing synthetic and real-world data. Extensive experiments show that the proposed method outperforms favorably against other state-of-the-art methods quantitatively and qualitatively. [ABSTRACT FROM AUTHOR]

Machine learning (ML) has automated a multitude of our day-to-day decision-making domains, such as education, employment, and driving automation. The continued success of ML largely depends on our ability to trust the model we are using. Recently, a new class of attacks called backdoor attacks have been developed. These attacks undermine the user’s trust in ML models. In this article, we present Neo, a model agnostic framework to detect and mitigate such backdoor attacks in image classification ML models. For a given image classification model, our approach analyzes the inputs it receives and determines if the model is backdoored. In addition to this feature, we also mitigate these attacks by determining the correct predictions of the poisoned images. We have implemented Neo and evaluated it against three state-of-the-art poisoned models. In our evaluation, we show that Neo can detect $\approx$ 88% of the poisoned inputs on average and it is as fast as 4.4 ms per input image. We also compare our Neo approach with the state-of-the-art defence methodologies proposed for backdoor attacks. Our evaluation reveals that despite being a blackbox approach, Neo is more effective in thwarting backdoor attacks than the existing techniques. Finally, we also reconstruct the exact poisoned input for the user to effectively test their systems. [ABSTRACT FROM AUTHOR]

Cameras with RGB-infrared (IR) image sensors are advantageous for applications that can benefit from simultaneous capture of color and near-IR (NIR) images. High quantum efficiency (QE) in NIR is desired for RGB-IR image sensors because it allows to reduce power of NIR illuminants, however, there is a tradeoff. As imaging with RGB-IR sensors is often done with visible/NIR dual bandpass filters, high NIR signal can lower the dynamic-range and the signal-to-noise ratio (SNR) of the color images. This work studies the effect of the NIR signal on the dynamic-range and the SNR of color and NIR images of RGB-IR image sensors. It presents a model that allows to evaluate these metrics using image sensor, camera, and scene parameters. The model is applied for the case of varied illuminant NIR content and for the case of varied pixel QE in NIR. Both cases are evaluated theoretically and experimentally, and experimental results are in good agreement with simulations. [ABSTRACT FROM AUTHOR]

Image and sentence matching has attracted much attention recently, and many effective methods have been proposed to deal with it. But even the current state-of-the-arts still cannot well associate those challenging pairs of images and sentences containing few-shot content in their regions and words. In fact, such a few-shot matching problem is seldom studied and has become a bottleneck for further performance improvement in real-world applications. In this work, we formulate this challenging problem as few-shot image and sentence matching, and accordingly propose an Aligned Cross-Modal Memory (ACMM) model to deal with it. The model can not only softly align few-shot regions and words in a weakly-supervised manner, but also persistently store and update cross-modal prototypical representations of few-shot classes as references, without using any groundtruth region-word correspondence. The model can also adaptively balance the relative importance between few-shot and common content in the image and sentence, which leads to better measurement of overall similarity. We perform extensive experiments in terms of both few-shot and conventional image and sentence matching, and demonstrate the effectiveness of the proposed model by achieving the state-of-the-art results on two public benchmark datasets. [ABSTRACT FROM AUTHOR]

Objective quality estimation of media content plays a vital role in a wide range of applications. Though numerous metrics exist for 2D images and videos, similar metrics are missing for 3D point clouds with unstructured and non-uniformly distributed points. In this paper, we propose ${\sf GraphSIM}$ GraphSIM —a metric to accurately and quantitatively predict the human perception of point cloud with superimposed geometry and color impairments. Human vision system is more sensitive to the high spatial-frequency components (e.g., contours and edges), and weighs local structural variations more than individual point intensities. Motivated by this fact, we use graph signal gradient as a quality index to evaluate point cloud distortions. Specifically, we first extract geometric keypoints by resampling the reference point cloud geometry information to form an object skeleton. Then, we construct local graphs centered at these keypoints for both reference and distorted point clouds. Next, we compute three moments of color gradients between centered keypoint and all other points in the same local graph for local significance similarity feature. Finally, we obtain similarity index by pooling the local graph significance across all color channels and averaging across all graphs. We evaluate ${\sf GraphSIM}$ GraphSIM on two large and independent point cloud assessment datasets that involve a wide range of impairments (e.g., re-sampling, compression, and additive noise). ${\sf GraphSIM}$ GraphSIM provides state-of-the-art performance for all distortions with noticeable gains in predicting the subjective mean opinion score (MOS) in comparison with point-wise distance-based metrics adopted in standardized reference software. Ablation studies further show that ${\sf GraphSIM}$ GraphSIM can be generalized to various scenarios with consistent performance by adjusting its key modules and parameters. Models and associated materials will be made available at https://njuvision.github.io/GraphSIM or http://smt.sjtu.edu.cn/papers/GraphSIM. [ABSTRACT FROM AUTHOR]

Objective: This translational study aims to investigate the clinical benefits of indocyanine green (ICG) based near-infrared window II (NIR-II) fluorescence image-guided surgery (FGS) on high-grade glioma (HGG) patients. Methods: Patients were randomly assigned to receive FGS or traditional white light image-guided surgery (WLS). The detection rate of NIR-II fluorescence was observed. Complete resection rate, progression-free survival (PFS), overall survival (OS), and neurological status were compared. Tissue samples were obtained from the FGS group, with the diagnosis based on the surgeons and the fluorescence recorded for comparison of diagnostic capability. Patients with WHO grade III gliomas or glioblastomas (GBM) were analyzed separately. Results: 15 GBM and 4 WHO grade III glioma patients in the FGS group and 18 GBM and 4 WHO grade III glioma patients in the WLS group were enrolled. The detection rate of NIR-II fluorescence was 100% for GBM. The complete resection rate was significantly increased by the FGS for GBM (FGS, 100% [95% CI 73.41-100] vs. WLS, 50% [95% CI 29.03-70.97], P = 0.0036). The PFS and OS of the FGS group were also significantly prolonged (Median PFS: FGS, 9.0 months vs. WLS, 7.0 months, P < 0.0001; Median OS: FGS, 19.0 months vs. WLS, 15.5 months, P = 0.0002). No recurrence was observed in WHO grade III glioma patients. Conclusions: NIR-II FGS achieves a much better complete resection rate of GBM than conventional WLS, leading to greatly improved survival of GBM patients. Significance: NIR-II FGS is a highly promising technique worthy of exploring more clinical applications. [ABSTRACT FROM AUTHOR]

With the increase in mobility in technology, artists are looking for ways to create paintings using their tablets and smartphones while on the move. Digital oil painting is a revolutionary technique where an artist uses digital paintbrush to simulate the traditional bristle paintbrush of a painting to get an oil paint image. In this article, we propose a novel realistic technique for reversible digital oil painting to accelerate the users to (re)paint an oil paint image to convey the surrounding atmosphere, express their emotions and feelings, or showcase a time of day or type of weather through that painting. Realism is all about accuracy. To repaint, the original image is retrieved from the existing oil paint image by reversing the pixels to its original value. We derive a mathematical expression to determine the probability of a pixel, with a set of neighbor pixels, being oil painted to get an oil paint image of the desired quality, which reduces the time and space complexities of the technique. We also develop prototypes on Android, Windows, and Ubuntu operating systems that successfully implement the proposed technique. [ABSTRACT FROM AUTHOR]

Background: Social media like YouTube have transformative effects on technical communication. Technical communication scholars have attended to the increasing use of social media personally, pedagogically, and professionally. Our stream of research focuses on YouTube videos for educational purposes within the various research avenues. Literature review: YouTube has become a viable platform for learning. YouTube educational videos have been studied from many different perspectives, yet research on engagement with YouTube educational videos is scarce, despite the importance of engagement in both learning and social media. Following extant research on YouTube educational video features, we probe the effects of visual cueing on social media engagement. Research question: How does visual cueing (anchors and intrinsic visual features) affect social media engagement with YouTube educational videos? Methodology: We sampled 196 YouTube educational videos on 28 physics and astronomy topics, and extracted visual cueing from the videos and social media engagement information from YouTube. Multiple linear regression analyses were conducted. Results: Our analyses show that intrinsic visual features (color contrast and visual complexity) are significantly related to social media engagement (involvement, intimacy, and interaction), while anchors (math equations and models) are not. Conclusion: Our study supports the empirical knowledge on social media engagement with YouTube educational videos and expands on the technical communication research for YouTube educational videos. In addition, this research contributes to the literature on engagement by extending its relevance to the social media learning environment. Finally, our study provides content creators with new video design insights that can be used to enhance social media engagement with YouTube educational videos. [ABSTRACT FROM AUTHOR]

Timely identification of the aging stage of the transformer can effectively ensure the safe operation of the transformer and prevent the aging fault of the transformer. This work proposes the aging diagnosis of transformer using the Raman spectroscopic recurrence plot combined with learning vector quantization (LVQ) neural network. Oil-paper insulation samples at different aging stages are obtained by accelerated thermal aging tests. Acquisition of Raman spectra of insulating oil is from 800 to 3000 cm−1 and converts to the Raman spectroscopic recurrence plot. The gray-level cooccurrence matrix (GLCM) is used to extract the characteristic information of the Raman spectroscopic recurrence plot. The optimal number of neurons in competitive layer of LVQ neural network is determined by the experimental method. A transformer aging diagnosis model based on the Raman spectroscopic recurrence plot is established. The results show that the diagnostic accuracy of the proposed diagnostic model reached 95%. The Raman spectroscopic recurrence plot has more abundant aging “fingerprint” characteristic information and the proposed diagnostic method shows potential for monitoring oil-paper insulation. [ABSTRACT FROM AUTHOR]

In Peru, the number of vehicles increased in the last decade, therefore, the automatic control requires a long database with a specific license plate model. It should be used for registration, evaluation and information extraction associated to the cars to control access parking with a particular license plate characteristic, especially for government buildings, therefore, an automatic evaluation of the license plate should be more dynamic and effective than European systems due to quantity of characters, specific segmentation and additional information in the Peruvian plate. License plate recognition is a widely applied system in artificial vision for the automatic obtaining of car license plates. This research article seeks to design a Peruvian license plate recognition system to reduce the time of vehicle registration and it involves accurate recognition of the plate location and extraction. Image processing techniques are used using the Python programming language, together with the OpenCV library. In addition, with YoloV4 a neural network is trained to locate the area where the license plate is located to facilitate the application of an Optical Character Recognizer (OCR). Our findings are a new improvement and evaluation in the traditional license plate software, with a new Peruvian database, so it allowed extracting the registration information instantly with high accuracy evaluated with 200 to 1000 images; therefore, the new contribution is the improvement in the false positive values with an accuracy of 100%, rate of failure of 0% and the sensibility of 100% with a specificity of 100% (neural network trained with 1000 images); besides it is the database for the future works for Peruvian cars. [ABSTRACT FROM AUTHOR]

Distance education has become an alternative in teaching derived from the Covid-19 pandemic. However, distance education has led to bad practices for some students. For example, it was detected that some students spoofed the teacher in a class or exam. Therefore, facial biometrics can be used to solve, in real-time, the spoofing problem. However, the solution is not exempt from presentation attacks that undermine the reliability of the systems. Other challenges that must be considered are lighting, resolution, and variable size of the faces, among others. In this paper, we present a methodology to address the problem of facial spoofing attacks. We combine the Extended Local Binary Patterns (ELBP) descriptor and YCbCr, HSV color models to highlight the saturation and illumination of an image. For the experiments, we present a comparison of our proposal against other state-of-the-art methods. We obtain an error of 2.45% with the Half Total Error Rate (HTER) metric in the MSU image bank. The results revealed that for environments where the camera resolutions are not controlled, our proposal provides a feasible solution reducing the costs of acquiring specific hardware [ABSTRACT FROM AUTHOR]

This paper presents a framework for low-light color imaging using a dual camera system that combines a high spatial resolution monochromatic (HSR-mono) image and a low spatial resolution color (LSR-color) image. We propose a cross-camera synthesis (CCS) module to learn and transfer illumination, color, and resolution attributes across paired HSR-mono and LSR-color images to recover brightness- and color-adjusted high spatial resolution color (HSR-color) images at both camera views. Jointly characterizing various attributes for final synthesis is extremely challenging because of significant domain gaps across cameras. Our proposed CCS method consists of three subtasks: reference-based illumination enhancement (RefIE), reference-based appearance transfer (RefAT), and reference-based super resolution (RefSR), by which we can characterize, transfer, and enhance illumination, color, and resolution at both views. Each subtask is implemented using deep neural networks (DNNs) that are first trained for each subtask separately and then finetuned jointly. Experimental results show the superior qualitative and quantitative results of the proposed CCS model on both synthetic content from popular datasets and real-captured scenes. Ablation studies further verify the model generalization to various exposures and camera baselines. We will make our work accessible at https://njuvision.github.io/CCS for reproducible research. [ABSTRACT FROM AUTHOR]

Tongue diagnosis is a novel and non-invasive approach commonly employed to carry out the supplementary diagnosis over the globe. Recently, several deep learning (DL) based tongue color image analysis models have existed in the literature for the effective detection of diseases. This paper presents a fusion of handcrafted with deep features based tongue color image analysis (FHDF-TCIA) technique to biomedical applications. The proposed FDHF-TCIA technique aims to investigate the tongue images using fusion model, and thereby determines the existence of disease. Primarily, the FHDF-TCIA technique comprises Gaussian filtering based preprocessing to eradicate the noise. The proposed FHDF-TCIA model encompasses a fusion of handcrafted local binary patterns (LBP) with MobileNet based deep features for the generation of optimal feature vectors. In addition, the political optimizer based quantum neural network (PO-QNN) based classification technique has been utilized for determining the proper class labels for it. A detailed simulation outcomes analysis of the FHDF-TCIA technique reported the higher accuracy of 0.992. [ABSTRACT FROM AUTHOR]

Distributed denial of service (DDoS) flooding attacks are very harmful and difficult to prevent due to the default-on nature of both inter-domain and intra-domain routing adopted by the current Internet. Accordingly, many future Internet initiatives try to eliminate DDoS attacks by design. Among these efforts, using path identifiers (PIDs) as inter-domain routing objects has attracted much research interest. However, existing approaches either advertise PIDs throughout the Internet or use secret but forgeable (whether static or dynamic) PIDs, which makes it easy to launch DDoS flooding attacks. To address this issue, in this paper we propose K-PID that uses cryptographic (thus unforgeable) PIDs as inter-domain routing objects. In particular, an ${N}$ -bit PID is comprised of an ${n}$ -bit prefix and (${N}\,\,-\,\,{n}$) bits that are cryptographic hash over per-flow/request information and a secret number. The prefix is used for inter-domain packet forwarding and the cryptographic hash is used as a token for allowing a data packet to enter into (or, pass through) a domain. We analyze K-PID’s performance in preventing DDoS flooding attacks and implement K-PID in a prototype to verify K-PID’s feasibility. The results show that K-PID can effectively prevent DDoS flooding attacks at a low cost of reducing the forwarding capability of routers. [ABSTRACT FROM AUTHOR]

Although person re-identification (person re-id) has advanced substantially in recent years, most methods are based on the assumption that the identities would not change clothes. This assumption might not hold in practice considering criminals who intentionally change clothes. In this work, we attempt to solve person re-id under moderate clothing change. Since the human body shape is considered as relatively more invariant under moderate clothing changes, we propose to learn a reliable shape-aware feature representation by mutually learning both colorful images and contour images. Instead of directly extracting shape features from contour images, we utilize contour feature learning as regularization and excavate more effective shape-aware feature representations from colorful images. We propose a multi-scale appearance and contour deep infomax (MAC-DIM) to maximize mutual information between colorful appearance features and contour shape features, and in this way, the extracted appearance features are constrained to be shape-aware in terms of both low-level visual properties and high-level semantics. To better model the long-range human body shape and explicitly capture contour segment relations, we introduce hierarchical graph modeling as aggregation headers, propagating structural context through graph convolutional networks (GCNs). The extensive results on benchmarks under clothing changes demonstrate the effectiveness of our shape-aware feature learning scheme. [ABSTRACT FROM AUTHOR]

Due to light scatter and absorption in waterbody, underwater imaging can be easily impaired with low contrast and visual distortion. The resulting images are often unable to meet the quality requirements of human perception and computer processing. Therefore, Underwater Image Enhancement (UIE) has been attracting extensive research efforts. Although deep learning has demonstrated its great success in many vision tasks, its huge amounts of parameters and computations are not conducive to UIE in resource-limited scenarios. In this paper, we address this issue by proposing a Lightweight Cascaded Network (LCNet) based on Laplacian image pyramids. At each pyramid level, we implement cascaded blocks upon a residual network. Specifically, high quality residuals can be progressively predicted with significantly reduced complexity in a coarse-to-fine fashion. Furthermore, these sub-networks are recursively nested to build our LCNet, thereby reducing the overall computational complexity with reused parameters. Extensive experiments demonstrate that the proposed method performs favorably against the state-of-the-arts in terms of visual quality, model parameters and complexity. [ABSTRACT FROM AUTHOR]

The rapidly growing demand for high-bandwidth data rate has pushed the wireline communications speed up to 112 Gb/s. In such high-speed and high-channel-count applications, a major challenge for differential pin map design is the tradeoff between cost and the crosstalk between differential pairs in the area of BGA pins and PCB via connection. In this letter, a pin map pattern with a high signal-to-ground ratio (S:G) of 1:1.25 and low crosstalk is proposed. Compared with the conventional pin map pattern in a square array with S:G = 1:1, the integrated crosstalk noise (ICN) is reduced by 75% and the insertion loss to crosstalk ratio (ICR) is increased by 13 dB. Compared with the conventional pin map pattern in a triangular array with S:G = 1:1.5, the ICN is reduced by 20%, and the ICR is improved by 3 dB. [ABSTRACT FROM AUTHOR]

In this study, a color detection phenomenon that does not utilize color filters is evaluated via the standard CMOS process. The device comprises multiple pn-junctions stacked in the depth direction, which enable division of the light responses by different light absorptions with respect to wavelength. To eliminate the color filters, we use rectifiers to cut the reverse currents corresponding to the long-wavelength infrared (Ir) regions. Numerical simulations predict the dynamic range of the sensor as 160 dB after removal of the filters that can decrease this value. Applications based on the Internet of Things can be realized via mass production at low cost by eliminating the additional processes required to form the conventional color filters in front of the sensor. We evaluate the proposed device by both numerical simulation and experiment as well as clarify that the sensing phenomenon requires sufficient PWell thickness to separate the depletion regions for successful operation. The appropriate process parameters for the CMOS process are revealed, providing useful information for testing under a wide range of process conditions. [ABSTRACT FROM AUTHOR]

In this paper, the image sensor compensation frameworks are established based on the correlation of pixel response characteristics; The pixel response model is established, and a method to measure the crosstalk of adjacent pixels of color image sensor under flat field light is proposed; at the same time, an artificial neural network training set is constructed by using the measured values and theoretical values of pixel response generated by combined exposures; a compensation method of using neural network compensation framework to replace high-dimensional neural network to traverse the image is proposed, which reduces the scale and training complexity of neural network; Finally, the corresponding spatial arrangement data of pixels are transformed into the frequency domain through the Fourier expansion algorithm, and the compensation effect is evaluated according to the change of high-frequency components. According to the experimental results, this method can effectively suppress color aliasing and crosstalk noise of the color image sensor. This paper provides a new method to compensate color aliasing and crosstalk noise. [ABSTRACT FROM AUTHOR]

Multiplexing anti-counterfeiting technologies based on micro/nano structural colors of photonic crystals have been proposed for information security because of unduplicability and diversity. However, most of current fabrication methods for micro/nano structural color such as electronic beam etching and ordered nanoparticles are accompanied with expensive and complex fabrication process. Reliable multiplexing anti-counterfeiting technologies with low cost and easy-fabrication process for currency, brands and information security remain a long-time and innovative challenge. Herein, a simple strategy to fabricate optical multiplexing anti-counterfeiting film based on self-assembled three-dimensional photonic crystal BPLC encoded with holography lithographic 1D/2D surface photonic crystal has been proposed. The proposed bilayer structure can exhibit multi-images with different angles in 3D space, leading to multiple encryptions with information protection or access permission. Quick response (QR) code with cryptographic functionality through our proposed multiplexing anti-counterfeiting film based on bilayer structural color is also demonstrated. The proposed strategy for optical multiplexing anti-counterfeiting film is low-cost, easy-fabricated, time saving, stable and flexible, capable for promising perspective in information security level and storage density, and offering potential application in anti-counterfeiting. [ABSTRACT FROM AUTHOR]

Compared with normal color vision observers (NOs), anomalous trichromats (AT) have anomalous color vision responses owing to differences in cone sensitivity. The wide-color-gamut display provides a promising solution to this problem. Herein, we propose a color discrimination experiment based on a wide-color-gamut display. The observers’ abilities to discern colors in various color directions were tested. The experimental findings were confirmed by simulations based on the shifting of the photopigments spectral absorption. It is concluded that the wide-color-gamut display significantly improves the color perception ability of AT, while proper light source optimization significantly enhances the visualization of AT. The study findings show that this AT-based, color vision model, may open up opportunities for future visual perception and biomedical applications. [ABSTRACT FROM AUTHOR]

Non-contact human body vital parameter measurements are getting more attention and have been extensively studied in a short span of time. One of such techniques is the heart rate (HR) estimation based on video recording of human face known as remote photoplethysmogram (rPPG). Usually, the recorded video gets contaminated due to illumination variation of ambient light, motion artifacts, and other environmental factors. Thus, extracting a reliable rPPG signal is a challenging task. In this paper, a novel spatial-temporal filtering method is proposed that utilizes 2D variational mode decomposition (2D-VMD) along with azimuthally averaged power spectrum density (AAPSD) and multimode kurtosis to extract a reliable rPPG signal. The robustness of the proposed algorithm is tested and validated using our own database and the publicly available standard dataset. The obtained experimental results are compared with the reference PPG signal measurements. Also, the proposed method is compared with the well-established independent component analysis (ICA)-based method. The performance results show that the non-contact HR estimated by the proposed method dramatically reduces the error and it proves our method to be superior than the existing method. [ABSTRACT FROM AUTHOR]

Robust traffic sign detection and recognition (TSDR) is of paramount importance for the successful realization of autonomous vehicle technology. The importance of this task has led to vast amount of research efforts and many promising methods have been proposed in the existing literature. However, most of these methods have been evaluated on clean and challenge-free datasets and overlooked the performance deterioration associated with different challenging conditions (CCs) that obscure the traffic-sign images captured in the wild. In this paper, we look at the TSDR problem under CCs and focus on the performance degradation associated with them. To this end, we propose a Convolutional Neural Network (CNN) based prior enhancement focused TSDR framework. Our modular approach consists of a CNN-based challenge classifier, Enhance-Net–an encoder-decoder CNN architecture for image enhancement, and two separate CNN architectures for sign-detection and classification. We propose a novel training pipeline for Enhance-Net that focuses on the enhancement of the traffic sign regions (instead of the whole image) in the challenging images subject to their accurate detection. We used CURE-TSD dataset consisting of traffic videos captured under different CCs to evaluate the efficacy of our approach. We experimentally show that our method obtains an overall precision and recall of 91.1% and 70.71% that is 7.58% and 35.90% improvement in precision and recall, respectively, compared to the current benchmark. Furthermore, we compare our approach with different CNN-based TSDR methods and show that our approach outperforms them by a large margin. [ABSTRACT FROM AUTHOR]

Modifying facial attributes without the paired dataset proves to be a challenging task. Previously, approaches either required supervision from a ground-truth transformed image or required training a separate model for mapping every pair of attributes. These limit the scalability of the models to accommodate a larger set of attributes since the number of models that we need to train grows exponentially large. Another major drawback of the previous approaches is the unintentional gain of the identity of the person as they transform the facial attributes. We propose a method that allows for controllable and identity-aware transformations across multiple facial attributes using only a single model. Our approach is to train a generative adversarial network (GAN) with a multitask conditional discriminator that recognizes the identity of the face, distinguishes real images from fake, as well as identifies facial attributes present in an image. This guides the generator into producing an output that is realistic while preserving the person’s identity and facial attributes. Through this framework, our model also learns meaningful image representations in a lower dimensional latent space and semantically associate separate parts of the encoded vector with both the person’s identity and facial attributes. This opens up the possibility of generating new faces and other transformations such as making the face thinner or chubbier. Furthermore, our model only encodes the image once and allows for multiple transformations using the encoded vector. This allows for faster transformations since it does not need to reprocess the entire image for every transformation. We show the effectiveness of our proposed method through both qualitative and quantitative evaluations, such as ablative studies, visual inspection, and face verification. Competitive results are achieved compared to the main competition (CycleGAN), however, at great space and extensibility gain by using a single model. [ABSTRACT FROM AUTHOR]

Multilabel learning, which handles instances associated with multiple labels, has attracted much attention in recent years. Many extant multilabel feature selection methods target global feature selection, which means feature selection weights for each label are shared by all instances. Also, many extant multilabel classification methods exploit global label selection, which means labels correlations are shared by all instances. In real-world objects, however, different subsets of instances may share different feature selection weights and different label correlations. In this article, we propose a novel framework with local feature selection and local label correlation, where we assume instances can be clustered into different groups, and the feature selection weights and label correlations can only be shared by instances in the same group. The proposed framework includes a group-specific feature selection process and a label-specific group selection process. The former process projects instances into different groups by extracting the instance-group correlation. The latter process selects labels for each instance based on its related groups by extracting the group-label correlation. In addition, we also exploit the intergroup correlation. These three kinds of group-based correlations are combined to perform effective multilabel classification. The experimental results on various datasets validate the effectiveness of our approach. [ABSTRACT FROM AUTHOR]

Existing domain adaptation (DA) methods generally assume that different domains have identical label space, and the training data are only sampled from a single domain. This unrealistic assumption is quite restricted for real-world applications, since it neglects the more practical scenario, where the source domain can contain the categories that are not shared by the target domain, and the training data can be collected from multiple modalities. In this article, we address a more difficult but practical problem, which recognizes RGB images through training on RGB-D data under the label space inequality scenario. There are three challenges in this task: 1) source and target domains are affected by the domain mismatch issue, which results in that the trained models perform imperfectly on the test data; 2) depth images are absent in the target domain (e.g., target images are captured by smartphones), when the source domain contains both the RGB and depth data. It makes the ordinary visual recognition approaches hardly applied to this task; and 3) in the real world, the source and target domains always have different numbers of categories, which would result in a negative transfer bottleneck being more prominent. Toward tackling the above challenges, we formulate a deep model, called visual-depth matching network (VDMN), where two new modules and a matching component can be trained in an end-to-end fashion jointly to identify the common and outlier categories effectively. The significance of VDMN is that it can take advantage of depth information and handle the domain distribution mismatch under label inequality simultaneously. The experimental results reveal that VDMN exceeds the state-of-the-art performance on various DA datasets, especially under the label inequality scenario. [ABSTRACT FROM AUTHOR]

This work is oriented toward video tracking of vehicles in a typical traffic environment, based on particle filters. The proposed tracking algorithm is based on simultaneous usage of three different image features - color, edge orientation, and texture. All three features are related to the contents of a rectangular window that includes both the vehicle that is tracked and local background and they are represented in the form of appropriate histograms. Based on individual estimates produced by every single feature, the resultant estimate is made by their weighted averaged. Weighting factors are adaptively changing depending on the quality of a particular feature, estimated by calculations of average similarities between the reference window and the set of windows on particles' positions. The tracking accuracies of single-feature and three-features-based filters have been verified using the set of traffic sequences illustrating the presence of typical disturbances (shadows, partial and full occlusions, maneuvering etc.).

Classifying pixels according to color, and segmenting the respective areas, are necessary steps in any computer vision task that involves color images. The gap between human color perception, linguistic color terminology, and digital representation are the main challenges for developing methods that properly classify pixels based on color. To address these challenges, we propose a novel method combining geometric analysis, color theory, fuzzy color theory, and multi-label systems for the automatic classification of pixels into 12 conventional color categories, and the subsequent accurate description of each of the detected colors. This method presents a robust, unsupervised, and unbiased strategy for color naming, based on statistics and color theory. The proposed model, “ABANICCO” (AB ANgular Illustrative Classification of COlor), was evaluated through different experiments: its color detection, classification, and naming performance were assessed against the standardized ISCC–NBS color system; its usefulness for image segmentation was tested against state-of-the-art methods. This empirical evaluation provided evidence of ABANICCO’s accuracy in color analysis, showing how our proposed model offers a standardized, reliable, and understandable alternative for color naming that is recognizable by both humans and machines. Hence, ABANICCO can serve as a foundation for successfully addressing a myriad of challenges in various areas of computer vision, such as region characterization, histopathology analysis, fire detection, product quality prediction, object description, and hyperspectral imaging.

This paper presents an in-depth analysis and research on the application of intelligent image color technology in teaching Chinese painting color. A feature reorganization-based approach to Chinese painting image style migration is proposed. Using the connection between depth features in neural networks and image semantics, the artistic style expression method of traditional Chinese painting is combined with existing methods to match the feature matrix in the generative network by feature reorganization, and finally, the migrated image is generated using a decoder. The texture of the target image is input to the generator as an additional condition, and the discriminator is changed to a relativistic discriminator to build a new generative adversarial network. To demonstrate the effectiveness of the method, a large number of Chinese ink and wash style landscape paintings and natural landscape photos are used as experimental data. The experimental results show that the obtained generated images after style migration are 8% better in realism and quality. An image color intelligent analysis system was designed, which integrates a color classification module, a pigment analysis module, an output saving module, and a pigment sample library to achieve the automatic classification of different color areas in Chinese painting, and ultimately allows a judgment to be made on the most likely pigment to be used for each type of color. In the teaching of Chinese painting, the system can transform the traditional "ear-listening" color transmission through vision into eye perception, which makes the teaching of Chinese painting vivid, arouses the enthusiasm of the classroom, and greatly promotes the color teaching of Chinese painting. [ABSTRACT FROM AUTHOR]

Image enhancement is a basic requirement for any computer vision application for further processing of an image. A common limitation with most of the existing methods, when applied to nearly invisible images, is the loss of color details during the enhancement process. So, a fuzzy c-means clustering-based method for image enhancement is proposed which enhances the perceptually invisible image along with preserving its color and naturalness. In this method, the image pixels are grouped into different clusters and are assigned membership values to those clusters. Based on this membership value, its intensity level is modified in the spatial domain. Modification of the gray levels proportional to the membership values leads to the stretching of the image histogram, similar in shape, to the original histogram. The process results in a very small shift in the mean intensity which preserves the color and brightness-related information of the image. The method enhances the image contrast and maintains the naturalness without introducing any artifacts. The simulation results on standard datasets reflect that the proposed algorithm is superior to many state-of-the-art and traditional methods for perceptually invisible images. [ABSTRACT FROM AUTHOR]

In recent years, learning-based single image dehazing networks have been comprehensively developed. However, performance improvement is limited due to domain shift between trained synthetic hazy images and untrained real-world hazy images. To alleviate this issue, this paper proposes a real-world dehazing targeted training scheme which nearly realizes paired real-world data training. As a result, a Twofold Multi-scale Generative Adversarial Network (TMS-GAN) consisting of a Haze-generation GAN (HgGAN) and a Haze-removal GAN (HrGAN) is designed. HgGAN attributes real haze properties to synthetic images and HrGAN removes haze from both synthetic and generated fake realistic data under supervision. Thus, the proposed method can better adapt to real-world image dehazing using this cooperative training scheme. Meanwhile, several structural advances of TMS-GAN also improve dehazing performance. Specifically, a haze residual map based on atmospheric scattering model is deduced in HgGAN for fake realistic data generation. The dual-branch generator in HrGAN draws attention to detail restoration by one branch along with another color-branch. A plug-and-play Multi-attention Progressive Fusion Module (MAPFM) is proposed and inserted in both HgGAN and HrGAN. MAPFM incorporates multi-attention mechanism to guide multi-scale feature fusion in a progressive manner, in which Adjacency-attention Block (AAB) can capture contributing features of each level and Self-attention Block (SAB) can establish non-local dependency of feature fusion. Experiments on mainstream benchmarks show that the proposed framework is superior especially on real-world hazy images among single image dehazing methods. [ABSTRACT FROM AUTHOR]