Numerous efforts have been made to design different low level saliency cues for the RGBD detection, such as color or depth contrast features, background and compactness priors. However, how these interact with each other incorporate effectively generate a master map remain challenging problem. In this paper, we new convolutional neural network (CNN) fuse into hierarchical features automatically detecting salient objects in images. existing works that directly feed raw image pixels CNN,...

Interpretation and diagnosis of machine learning models have gained renewed interest in recent years with breakthroughs new approaches. We present Manifold, a framework that utilizes visual analysis techniques to support interpretation, debugging, comparison more transparent interactive manner. Conventional usually focus on visualizing the internal logic specific model type (i.e., deep neural networks), lacking ability extend complex scenario where different types are integrated. To this...

Although microplastics have been detected in human blood, placenta and other tissues. In this study, for the first time, we characterized presence variation of microplastic deposition patterns three skeletal tissues, namely bone, cartilage, intervertebral discs. Forty fragments were observed 24 samples from disc, ranging 25.44 to 407.39 μm diameter. The abundance disc (61.1 ± 44.2 particles/g) was higher than those bone (22.9 15.7 cartilage tissue (26.4 17.6 particles/g). average sizes discs...

3D hand pose tracking/estimation will be very important in the next generation of human-computer interaction. Most currently available algorithms rely on low-cost active depth sensors. However, these sensors can easily interfered by other sources and require relatively high power consumption. As a result, they are not suitable for outdoor environments mobile devices. This paper aims at tracking/estimating poses using passive stereo which avoids limitations. A benchmark with 18,000 image...

Person re-identification (ReID) has gained an impressive progress in recent years. However, the occlusion is still a common and challenging problem for ReID methods. Several mainstream methods utilize extra cues (e.g., human pose information) to distinguish parts from obstacles alleviate problem. Although achieving inspiring progress, these severely rely on fine-grained cues, are sensitive estimation error cues. In this paper, we show that existing may degrade if information sparse or noisy....

Despite recent stereo matching networks achieving impressive performance given sufficient training data, they suffer from domain shifts and generalize poorly to unseen domains. We argue that maintaining feature consistency between pixels is a vital factor for promoting the generalization capability of networks, which has not been adequately considered. Here we address this issue by proposing simple pixel-wise contrastive learning across viewpoints. The loss function explicitly constrains...

In this paper we establish a long-term 3D hand pose tracking benchmark <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> . It contains 18,000 stereo image pairs as well the ground-truth positions of palm and finger joints from different scenarios. Meanwhile, to accurately segment images, propose novel stereo-based segmentation depth estimation algorithm specially tailored for here. The experiments indicate effectiveness proposed by...

This paper presents an efficient visualization and exploration approach for modeling characterizing the relationships uncertainties in context of a multidimensional ensemble dataset. Its core is novel dissimilarity-preserving projection technique that characterizes not only among mean values data objects but also distributions members. uncertainty-aware scheme leads to improved understanding intrinsic structure The analysis dataset further augmented by suite visual encoding tools....

This paper presents ER-NeRF, a novel conditional Neural Radiance Fields (NeRF) based architecture for talking portrait synthesis that can concurrently achieve fast convergence, real-time rendering, and state-of-the-art performance with small model size. Our idea is to explicitly exploit the unequal contribution of spatial regions guide modeling. Specifically, improve accuracy dynamic head reconstruction, compact expressive NeRF-based Tri-Plane Hash Representation introduced by pruning empty...

In this paper, we propose an efficient algorithm to directly restore a clear image from hazy input. The proposed hinges on end-to-end trainable neural network that consists of encoder and decoder. is exploited capture the context derived input images, while decoder employed estimate contribution each final dehazed result using learned representations attributed encoder. constructed adopts novel fusion-based strategy which derives three inputs original by applying White Balance (WB), Contrast...

Effective training of deep neural networks (DNNs) usually requires labeling a large dataset, which is time and labor intensive. Recently, various data augmentation strategies like regional dropout mix have been proposed, are effective as the augmented dataset can guide model to attend on less discriminative parts. However, these operate only at image level, where objects background coupled. Thus, boundaries not well due fixed semantic scenario. In this paper, we propose ObjectAug perform...

Learning RAW-to-sRGB mapping has drawn increasing attention in recent years, wherein an input raw image is trained to imitate the target sRGB captured by another camera. However, severe color inconsistency makes it very challenging generate well-aligned training pairs of and images. While learning with inaccurately aligned supervision prone causing pixel shift producing blurry results. In this paper, we circumvent such issue presenting a joint model for alignment mapping. To diminish effect...

Visualizations in organizational research have primarily been used the context of traditional survey data, where individual data points (e.g., responses) can typically be plotted, and qualitative language data) quantitative frequency information are not combined. Moreover, visualizations a hypothetico-deductive fashion to showcase significant hypothesized results. With advent big which has characterized as being particularly high volume, variety, velocity collection, need more explicitly...

Abstract Real‐time microblogs can be utilized to provide situational awareness during emergency and disaster events. However, the utilization of these datasets requires decision makers perform their exploration analysis across a range data scales from local global, while maintaining cohesive thematic context transition between different granularity levels. The information dimensions at varied human remains non‐trivial task. To this end, we present visual analytics environment that supports...

Exemplar-based face sketch synthesis methods usually meet the challenging problem that input photos are captured in different lighting conditions from training photos. The critical step causing failure is search of similar patch candidates for an photo patch. Conventional illumination invariant distances adopted rather than directly relying on pixel intensity difference, but they will fail when local contrast within a changes. In this paper, we propose fast preprocessing method named...

Automatically selecting exposure bracketing (images exposed differently) is important to obtain a high dynamic range image by using multi-exposure fusion. Unlike previous methods that have many restrictions such as requiring camera response function, sensor noise model, and stream of preview images with different exposures (not accessible in some scenarios e.g. mobile applications), we propose novel deep neural network automatically select bracketing, named EBSNet, which sufficiently...

Understanding of human vision system (HVS) has inspired many computer algorithms. Stereo matching, which borrows the idea from stereopsis, been extensively studied in existing literature. However, scant attention drawn on a typical scenario where binocular inputs are qualitatively different (e.g., high-res master camera and low-res slave dual-lens module). Recent advances optometry reveal capability visual to maintain coarse stereopsis under such visually imbalanced conditions. Bionically...

With the advance of technology, entities can be observed in multiple views. Multiple views containing different types features used for clustering. Although multi-view clustering has been successfully applied many applications, previous methods usually assume complete instance mapping between In real-world information gathered from sources, while each source contain views, which are more cohesive learning. The under same fully mapped, but they very heterogeneous. Moreover, mappings sources...

Human motion prediction is a classical problem in computer vision and graphics, which has wide range of practical applications. Previous effects achieve great empirical performance based on an encoding-decoding style. The methods this style work by first encoding previous motions to latent representations then decoding the into predicted motions. However, practice, they are still unsatisfactory due several issues, including complicated loss constraints, cumbersome training processes, scarce...