Visual Computing for Industry, Biomedicine, and Art最新文献

This paper presents a novel algorithm for planar G1 interpolation using typical curves with monotonic curvature. The G1 interpolation problem is converted into a system of nonlinear equations and sufficient conditions are provided to check whether there is a solution. The proposed algorithm was applied to a curve completion task. The main advantages of the proposed method are its simple construction, compatibility with NURBS, and monotonic curvature.

Data visualization blends art and science to convey stories from data via graphical representations. Considering different problems, applications, requirements, and design goals, it is challenging to combine these two components at their full force. While the art component involves creating visually appealing and easily interpreted graphics for users, the science component requires accurate representations of a large amount of input data. With a lack of the science component, visualization cannot serve its role of creating correct representations of the actual data, thus leading to wrong perception, interpretation, and decision. It might be even worse if incorrect visual representations were intentionally produced to deceive the viewers. To address common pitfalls in graphical representations, this paper focuses on identifying and understanding the root causes of misinformation in graphical representations. We reviewed the misleading data visualization examples in the scientific publications collected from indexing databases and then projected them onto the fundamental units of visual communication such as color, shape, size, and spatial orientation. Moreover, a text mining technique was applied to extract practical insights from common visualization pitfalls. Cochran's Q test and McNemar's test were conducted to examine if there is any difference in the proportions of common errors among color, shape, size, and spatial orientation. The findings showed that the pie chart is the most misused graphical representation, and size is the most critical issue. It was also observed that there were statistically significant differences in the proportion of errors among color, shape, size, and spatial orientation.

One common way to aid coaching and seek to improve athletes' performance is by recording training sessions for posterior analysis. In the case of sailing, coaches record videos from another boat, but usually rely on handheld devices, which may lead to issues with the footage and missing important moments. On the other hand, by autonomously recording the entire session with a fixed camera, the analysis becomes challenging owing to the length of the video and possible stabilization issues. In this work, we aim to facilitate the analysis of such full-session videos by automatically extracting maneuvers and providing a visualization framework to readily locate interesting moments. Moreover, we address issues related to image stability. Finally, an evaluation of the framework points to the benefits of video stabilization in this scenario and an appropriate accuracy of the maneuver detection method.

Acral melanoma (AM) is a rare and lethal type of skin cancer. It can be diagnosed by expert dermatologists, using dermoscopic imaging. It is challenging for dermatologists to diagnose melanoma because of the very minor differences between melanoma and non-melanoma cancers. Most of the research on skin cancer diagnosis is related to the binary classification of lesions into melanoma and non-melanoma. However, to date, limited research has been conducted on the classification of melanoma subtypes. The current study investigated the effectiveness of dermoscopy and deep learning in classifying melanoma subtypes, such as, AM. In this study, we present a novel deep learning model, developed to classify skin cancer. We utilized a dermoscopic image dataset from the Yonsei University Health System South Korea for the classification of skin lesions. Various image processing and data augmentation techniques have been applied to develop a robust automated system for AM detection. Our custom-built model is a seven-layered deep convolutional network that was trained from scratch. Additionally, transfer learning was utilized to compare the performance of our model, where AlexNet and ResNet-18 were modified, fine-tuned, and trained on the same dataset. We achieved improved results from our proposed model with an accuracy of more than 90 % for AM and benign nevus, respectively. Additionally, using the transfer learning approach, we achieved an average accuracy of nearly 97 %, which is comparable to that of state-of-the-art methods. From our analysis and results, we found that our model performed well and was able to effectively classify skin cancer. Our results show that the proposed system can be used by dermatologists in the clinical decision-making process for the early diagnosis of AM.

In this paper, we introduce a visual analytics approach aimed at helping machine learning experts analyze the hidden states of layers in recurrent neural networks. Our technique allows the user to interactively inspect how hidden states store and process information throughout the feeding of an input sequence into the network. The technique can help answer questions, such as which parts of the input data have a higher impact on the prediction and how the model correlates each hidden state configuration with a certain output. Our visual analytics approach comprises several components: First, our input visualization shows the input sequence and how it relates to the output (using color coding). In addition, hidden states are visualized through a nonlinear projection into a 2-D visualization space using t-distributed stochastic neighbor embedding to understand the shape of the space of the hidden states. Trajectories are also employed to show the details of the evolution of the hidden state configurations. Finally, a time-multi-class heatmap matrix visualizes the evolution of the expected predictions for multi-class classifiers, and a histogram indicates the distances between the hidden states within the original space. The different visualizations are shown simultaneously in multiple views and support brushing-and-linking to facilitate the analysis of the classifications and debugging for misclassified input sequences. To demonstrate the capability of our approach, we discuss two typical use cases for long short-term memory models applied to two widely used natural language processing datasets.

The visualization of dynamic graphs is a challenging task owing to the various properties of the underlying relational data and the additional time-varying property. For sparse and small graphs, the most efficient approach to such visualization is node-link diagrams, whereas for dense graphs with attached data, adjacency matrices might be the better choice. Because graphs can contain both properties, being globally sparse and locally dense, a combination of several visual metaphors as well as static and dynamic visualizations is beneficial. In this paper, a visually and algorithmically scalable approach that provides views and perspectives on graphs as interactively linked node-link and adjacency matrix visualizations is described. As the novelty of this technique, insights such as clusters or anomalies from one or several combined views can be used to influence the layout or reordering of the other views. Moreover, the importance of nodes and node groups can be detected, computed, and visualized by considering several layout and reordering properties in combination as well as different edge properties for the same set of nodes. As an additional feature set, an automatic identification of groups, clusters, and outliers is provided over time, and based on the visual outcome of the node-link and matrix visualizations, the repertoire of the supported layout and matrix reordering techniques is extended, and more interaction techniques are provided when considering the dynamics of the graph data. Finally, a small user experiment was conducted to investigate the usability of the proposed approach. The usefulness of the proposed tool is illustrated by applying it to a graph dataset, such as e co-authorships, co-citations, and a Comprehensible Perl Archive Network distribution.

As a disaster prevention measure based on self-assistance and mutual assistance, disaster prevention maps are being created with citizen participation throughout Japan. The process of creating disaster prevention maps is itself a disaster prevention measure that contributes to raising awareness of disaster prevention by promoting exchange and cooperation within the region. By focusing on relations between road networks and hazardous elements, we developed a system to support disaster prevention map creation that visualizes roads at high risk during a disaster and facilitates the study of evacuation simulations. This system leads to a completed disaster prevention map in three phases. In the first phase, we use a device with GPS logging functions to collect information related to hazardous elements. In the second phase, we use Google Maps ("online map," below) to visualize roads with high evacuation risk. In the final phase, we perform a regional evaluation through simulations of disaster-time evacuations. In experimental verifications, by conducting usability tests after creating a disaster prevention map in the target area, we evaluated the system in terms of simple operability and visibility. We found that by implementing this series of processes, even users lacking specialized knowledge regarding disaster prevention can intuitively discover evacuation routes while considering the relations between visualized road networks and hazardous elements. These results show that compared with disaster prevention maps having simple site notations using existing WebGIS systems, disaster prevention maps created by residents while inspecting the target area raise awareness of risks present in the immediate vicinity even in normal times and are an effective support system for prompt disaster prevention measures and evacuation drills.

To minimize radiation risk, dose reduction is important in the diagnostic and therapeutic applications of computed tomography (CT). However, image noise degrades image quality owing to the reduced X-ray dose and a possible unacceptably reduced diagnostic performance. Deep learning approaches with convolutional neural networks (CNNs) have been proposed for natural image denoising; however, these approaches might introduce image blurring or loss of original gradients. The aim of this study was to compare the dose-dependent properties of a CNN-based denoising method for low-dose CT with those of other noise-reduction methods on unique CT noise-simulation images. To simulate a low-dose CT image, a Poisson noise distribution was introduced to normal-dose images while convoluting the CT unit-specific modulation transfer function. An abdominal CT of 100 images obtained from a public database was adopted, and simulated dose-reduction images were created from the original dose at equal 10-step dose-reduction intervals with a final dose of 1/100. These images were denoised using the denoising network structure of CNN (DnCNN) as the general CNN model and for transfer learning. To evaluate the image quality, image similarities determined by the structural similarity index (SSIM) and peak signal-to-noise ratio (PSNR) were calculated for the denoised images. Significantly better denoising, in terms of SSIM and PSNR, was achieved by the DnCNN than by other image denoising methods, especially at the ultra-low-dose levels used to generate the 10% and 5% dose-equivalent images. Moreover, the developed CNN model can eliminate noise and maintain image sharpness at these dose levels and improve SSIM by approximately 10% from that of the original method. In contrast, under small dose-reduction conditions, this model also led to excessive smoothing of the images. In quantitative evaluations, the CNN denoising method improved the low-dose CT and prevented over-smoothing by tailoring the CNN model.

In recent years, simultaneous localization and mapping in dynamic environments (dynamic SLAM) has attracted significant attention from both academia and industry. Some pioneering work on this technique has expanded the potential of robotic applications. Compared to standard SLAM under the static world assumption, dynamic SLAM divides features into static and dynamic categories and leverages each type of feature properly. Therefore, dynamic SLAM can provide more robust localization for intelligent robots that operate in complex dynamic environments. Additionally, to meet the demands of some high-level tasks, dynamic SLAM can be integrated with multiple object tracking. This article presents a survey on dynamic SLAM from the perspective of feature choices. A discussion of the advantages and disadvantages of different visual features is provided in this article.

Scholarly communication of knowledge is predominantly document-based in digital repositories, and researchers find it tedious to automatically capture and process the semantics among related articles. Despite the present digital era of big data, there is a lack of visual representations of the knowledge present in scholarly articles, and a time-saving approach for a literature search and visual navigation is warranted. The majority of knowledge display tools cannot cope with current big data trends and pose limitations in meeting the requirements of automatic knowledge representation, storage, and dynamic visualization. To address this limitation, the main aim of this paper is to model the visualization of unstructured data and explore the feasibility of achieving visual navigation for researchers to gain insight into the knowledge hidden in scientific articles of digital repositories. Contemporary topics of research and practice, including modifiable risk factors leading to a dramatic increase in Alzheimer's disease and other forms of dementia, warrant deeper insight into the evidence-based knowledge available in the literature. The goal is to provide researchers with a visual-based easy traversal through a digital repository of research articles. This paper takes the first step in proposing a novel integrated model using knowledge maps and next-generation graph datastores to achieve a semantic visualization with domain-specific knowledge, such as dementia risk factors. The model facilitates a deep conceptual understanding of the literature by automatically establishing visual relationships among the extracted knowledge from the big data resources of research articles. It also serves as an automated tool for a visual navigation through the knowledge repository for faster identification of dementia risk factors reported in scholarly articles. Further, it facilitates a semantic visualization and domain-specific knowledge discovery from a large digital repository and their associations. In this study, the implementation of the proposed model in the Neo4j graph data repository, along with the results achieved, is presented as a proof of concept. Using scholarly research articles on dementia risk factors as a case study, automatic knowledge extraction, storage, intelligent search, and visual navigation are illustrated. The implementation of contextual knowledge and its relationship for a visual exploration by researchers show promising results in the knowledge discovery of dementia risk factors. Overall, this study demonstrates the significance of a semantic visualization with the effective use of knowledge maps and paves the way for extending visual modeling capabilities in the future.