Visual Framework Research Articles

Automatic pixel-level detection of tire defects based on a lightweight Transformer architecture

Efficient and accurate visual inspection of products is of great significance for intelligent manufacturing. In this paper, a visual inspection framework based on a lightweight Transformer is proposed for the pixel-level inspection of tire defects. A dual-path-Transformer feature encoder was proposed based on a hybrid architecture of convolutional neural network and Transformer, which is used to learn the local and global relationship of defect features. Moreover, multi-scale fusion Transformer (MFT) and spatial cross Transformer (SCT) were proposed, based on which a feature decoder was built. In the decoder, the MFT provides valuable spatial information for SCT, such that different levels of feature maps can refine the dependency of pixels through SCT. The proposed method is tested on the tire radiographic image dataset. Experimental results show that the proposed method reaches a detection accuracy of 98.57% and mIoU of 85.56%. Moreover, the method achieves the balance between accuracy and efficiency with satisfactory characterization ability for the defect geometric shape, and provides theoretical support for the future industrial deployment of Transformer.

Cover Image, Volume 38, Issue 9

Computer-Aided Civil and Infrastructure EngineeringVolume 38, Issue 9 COVERFree Access Cover Image, Volume 38, Issue 9 First published: 14 May 2023 https://doi.org/10.1111/mice.13044AboutPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Graphical Abstract On the cover: The cover image is based on the Research Article Visual comfort generative design framework based on parametric network in underground space by Yingbin Gui et al., https://doi.org/10.1111/mice.12936. Volume38, Issue9June 2023 RelatedInformation

The Fusion of Ancient and Contemporary: A Visual Rhetoric Analysis of the Visual Representation of Shaanxi in the Opening and Closing Ceremonies of the National Games

In recent years, the improvement of city image has become an important topic of concern, at the same time, the hosting of mega-events has become an indispensable part of building city image. Therefore, the impact of mega-events on city image has attracted widespread attention in the academic community. However, the current analysis of the impact of mega-events on city image improvement from the perspective of visual rhetoric has not received much attention. Thus, against this background, this study focuses on the visual turn of shaping city image in the Internet era, using the opening and closing ceremonies of the Xi'an National Games as an example. The author has coded the Shaanxi elements in the opening and closing ceremonies of the National Games and created a chart using the visual grammar framework. The study shows that Shaanxi elements can be classified into four types: signs, buildings, natural landscape and distinctive cultures. Based on the interactive meaning in visual grammar analysis, the author analyzes the presentation, function, and meaning of Shaanxi elements from two aspects: the "social distance" represented by lens language and the "attitude" represented by viewing angle. These Shaanxi elements ultimately built an image of Shaanxi with strong economic and cultural strength, distinctive characteristics, inheritance, and innovation. The image of Shaanxi spread by the National Games has become a vivid memory of a generation, and the construction of the image has also become a manifestation of the enhancement of Shaanxi's international communication ability.

Accurate delineation of individual tree crowns in tropical forests from aerial RGB imagery using Mask R‐CNN

AbstractTropical forests are a major component of the global carbon cycle and home to two‐thirds of terrestrial species. Upper‐canopy trees store the majority of forest carbon and can be vulnerable to drought events and storms. Monitoring their growth and mortality is essential to understanding forest resilience to climate change, but in the context of forest carbon storage, large trees are underrepresented in traditional field surveys, so estimates are poorly constrained. Aerial photographs provide spectral and textural information to discriminate between tree crowns in diverse, complex tropical canopies, potentially opening the door to landscape monitoring of large trees. Here we describe a new deep convolutional neural network method, Detectree2, which builds on the Mask R‐CNN computer vision framework to recognize the irregular edges of individual tree crowns from airborne RGB imagery. We trained and evaluated this model with 3797 manually delineated tree crowns at three sites in Malaysian Borneo and one site in French Guiana. As an example application, we combined the delineations with repeat lidar surveys (taken between 3 and 6 years apart) of the four sites to estimate the growth and mortality of upper‐canopy trees. Detectree2 delineated 65 000 upper‐canopy trees across 14 km2 of aerial images. The skill of the automatic method in delineating unseen test trees was good (F1 score = 0.64) and for the tallest category of trees was excellent (F1 score = 0.74). As predicted from previous field studies, we found that growth rate declined with tree height and tall trees had higher mortality rates than intermediate‐size trees. Our approach demonstrates that deep learning methods can automatically segment trees in widely accessible RGB imagery. This tool (provided as an open‐source Python package) has many potential applications in forest ecology and conservation, from estimating carbon stocks to monitoring forest phenology and restoration.Python package available to install at https://github.com/PatBall1/Detectree2.

ChroKit: a Shiny-based framework for interactive analysis, visualization and integration of genomic data.

We developed ChroKit (the Chromatin toolKit), an interactive web-based framework written in R that enables intuitive exploration, multidimensional analyses, and visualization of genomic data from ChIP-Seq, DNAse-Seq or any other NGS experiment that reports the enrichment of aligned reads over genomic regions. This program takes preprocessed NGS data and performs operations on genomic regions of interest, including resetting their boundaries, their annotation based on proximity to genomic features, the association to gene ontologies, and signal enrichment calculations. Genomic regions can be further refined or subsetted by user-defined logical operations and unsupervised classification algorithms. ChroKit generates a full range of plots that are easily manipulated by point and click operations, thus allowing 'on the fly' re-analysis and fast exploration of the data. Working sessions can be exported for reproducibility, accountability, and easy sharing within the bioinformatics community. ChroKit is multiplatform and can be deployed on a server to enhance computational speed and provide simultaneous access by multiple users. ChroKit is a fast and intuitive genomic analysis tool suited for a wide range of users due to its architecture and its user-friendly graphical interface. ChroKit source code is available at https://github.com/ocroci/ChroKit and the Docker image at https://hub.docker.com/r/ocroci/chrokit.

Visual localization with a monocular camera for unmanned aerial vehicle based on landmark detection and tracking using YOLOv5 and DeepSORT

Absolute visual localization is of significant importance for unmanned aerial vehicles when the satellite-based localization system is not available. With the rapid evolution in the field of deep learning, the real-time visual detection and tracking of landmarks by an unmanned aerial vehicle could be implemented onboard. This study demonstrates a landmark-based visual localization framework for unmanned aerial vehicles flying at low altitudes. YOLOv5 and DeepSORT are used for multi-object detection and tracking, respectively. The unmanned aerial vehicle localization is achieved according to the geometric similarity between the geotagged transmission towers and the annotated images captured by a monocular camera. The validation is accomplished both in the Rflysim-based simulation and the quadrotor-based real flight. The localization precision is about 10 m, and the location update frequency reaches 5 Hz with a commercially available entry-level edge artificial intelligence platform. The proposed visual localization strategy needs no satellite image as a reference map, which saves a significant amount of the GPU memory and makes possible the end-to-end implementation on small unmanned aerial vehicles.

Pembelajaran Mendalam Pengklasifikasi Ekspresi Wajah Manusia dengan Model Arsitektur Xception pada Metode Convolutional Neural Network

Deep learning is a neural network that creates innovations that give computer-implanted problem-solving expertise. One of the principles of computer vision is a detection system with a vision framework that can identify things encountered in the same manner as a human vision system. Using an artificial intelligence-based Convolutional Neural Network (CNN) model with deep learning techniques, we present a face emotion identification system. The categorization of facial expressions will be utilized as the basis for a face recognition system trained using CNN. The applications are intended to use the OpenCV, Keras, and TensorFlow libraries as the backend. We were discussing the study on the best use of xception architectural models in facial expression recognition systems. Based on the results of these tests, the study obtained an increased accuracy value in training and data testing on an xception architecture model trained for facial expressions using the FER-2013 dataset, resulting in an accuracy value of 66% as well as the value of each average for precision (76%), recall (65%), and F1 score (63%).

UD‐SV: Urban data services and visualization framework for sharing multidisciplinary research

AbstractWith the widespread availability of a large volume of urban data, stakeholders from different domains require advanced tools to manage, visualize and understand cities and their evolution. During the last few years, researchers have proposed numerous research works and applications to illustrate the cities of the past and possible scenarios of the future under different conditions. However, many of these approaches are one‐time solutions and not based on standards, making them obsolete and unusable for reproducible research. In this article, we present UD‐SV: an Urban data‐Services and Visualization open‐source framework for multidisciplinary research to handle complex processing, analysis, and visualization of urban data. However, our goal is not to present a one‐time monolithic software solution for urban data management and analysis, but we demonstrate the design and development of an open and interoperable software framework driven by use cases from diverse users to solve applied research challenges. The main contribution of UD‐SV is that it uses open standards and open data with documented and reproducible processes with a particular emphasis on the reuse of existing open‐source software components. We also show an enhanced use of standards to enable a shift toward components that are interchangeable or composable with other existing components in the GIS community.

Preliminary Investigation into the Affordances of Evaluative Resources in Multimodal Texts of an Established Textbook Series in Vietnam

The paper aims to initially examine the deployment of Attitude in multimodal reading texts in Solutions (Falla & Davies, 2013), an English textbook series being used as textbooks in most universities in Vietnam. To reach its aim, it first deployed the Appraisal framework proposed by Martin and White (2005) to examine evaluative resources in the fifteen verbal texts of the five genres of Anecdote, Biographical recount, Exposition, Narrative, and Review (Rose & Martin, 2012). A closer examination of how images contribute to articulating evaluation is then conducted on five of these multimodal reading texts in light of the Visual Appraisal frameworks suggested by Economou (2009) and White (2014) and the terms Convergence and Divergence proposed by Painter, Martin and Unsworth (2013). It was found that various aspects of Attitude are exploited with a range of frequency in different genres. It was also indicated that Convergence tends to be more frequently deployed than Divergence. In the same multimodal text, both cases of Convergence and Divergence may occur or only one of the two types. In any way, they all have the function of co-articulating attitude. The paper concludes by highlighting the necessity of selecting appropriate images in textbooks among editors and deploying images in effectively teaching multimodal texts among lecturers and teachers.

A neural network framework for immediate temperature prediction of surgical hand-held drilling

Background and objectiveHeat generation and associated temperature rise in surgical drilling can cause irreversible tissue damage. It is nearly impossible to provide immediate temperature prediction for a hand-held drilling process since both feed rate and motion vary with time. The objective of this study is to present and test a framework for immediate bone drilling temperature visualization based on a neural network (NN) model and a linear time-invariant (LTI) model. MethodsIn this study, the finite element analysis (FEA) model is used as the ground truth. The NN model is used to predict the location-dependent thermal responses of FEA, while LTI is used to superimpose these responses based on the location history of the heat source. The use of LTI can eliminate the uncertainty of the unlimited possibility in the time domain. To test the framework, two three-dimensional drilling cases are studied, one with a constant drilling feed and straight path and the other with a varying feed and a varying path. ResultsThe NN model using U-net architecture can achieve the predicted correlation of over 97% with only 1% of the total number of data points. Using the framework with U-net and LTI, both case studies show good agreement in temporal and spatial temperature distributions with the ground truth. The average error near the drilling path is less than 10%. Discrepancies are mainly found near the heat source and the regions near the removed material. ConclusionsAn FEA surrogate model for rapid and accurate prediction of 3D temperature during arbitrary bone drilling is successfully made. The overall error is less than 5% on average in the two case studies. Future improvements include strategies for training data selection and data formating.

Design of moving average chart and auxiliary information based chart using extended EWMA

The control chart is the most valuable tool in the manufacturing process to track the output process in the industries. Quality specialists always want a visual framework that recognizes sustainable improvements in the monitoring processes. The efficiency of a control chart is increased by utilizing a memory-based estimator or by using any extra information relevant to the key variable. In this study, we present Extended EWMA (EEWMA) and EWMA based monitoring charts for observing the process location using moving average (MA) statistic under two different situations, i.e., when some extra information is known and unknown. We also propose an EEWMA control chart using Auxiliary Information. The output of these charts is evaluated and contrasted to the various existing charts on the basis of average run length (ARL). The comparison indicates that the proposed charts outperform rivals in identifying all types of shifts in the process location parameter. The implementation of these plans is also rendered to incorporate them in a practical situation.

An open-access T-BAS phylogeny for emerging Phytophthora species.

Phytophthora species cause severe diseases on food, forest, and ornamental crops. Since the genus was described in 1876, it has expanded to comprise over 190 formally described species. There is a need for an open access phylogenetic tool that centralizes diverse streams of sequence data and metadata to facilitate research and identification of Phytophthora species. We used the Tree-Based Alignment Selector Toolkit (T-BAS) to develop a phylogeny of 192 formally described species and 33 informal taxa in the genus Phytophthora using sequences of eight nuclear genes. The phylogenetic tree was inferred using the RAxML maximum likelihood program. A search engine was also developed to identify microsatellite genotypes of P. infestans based on genetic distance to known lineages. The T-BAS tool provides a visualization framework allowing users to place unknown isolates on a curated phylogeny of all Phytophthora species. Critically, the tree can be updated in real-time as new species are described. The tool contains metadata including clade, host species, substrate, sexual characteristics, distribution, and reference literature, which can be visualized on the tree and downloaded for other uses. This phylogenetic resource will allow data sharing among research groups and the database will enable the global Phytophthora community to upload sequences and determine the phylogenetic placement of an isolate within the larger phylogeny and to download sequence data and metadata. The database will be curated by a community of Phytophthora researchers and housed on the T-BAS web portal in the Center for Integrated Fungal Research at NC State. The T-BAS web tool can be leveraged to create similar metadata enhanced phylogenies for other Oomycete, bacterial or fungal pathogens.

Democracy on the walls: the portrayal of the common people in post-dictatorship Iberian murals

ABSTRACT After the fall of the right-wing Estado Novo in Portugal in 1974 and the death of the military dictator Francisco Franco in Spain in 1975, many protest murals appeared in both countries on the walls of abandoned buildings and in marginalized urban spaces. These ephemeral grassroots artworks celebrated the collective liberation of the Iberian people from the cultural constraints of Western Europe’s two most enduring dictatorships. Murals constituted a striking visual response to the sociopolitical changes occurring in two countries that were experiencing the transition to democracy in very different ways. In this article, I argue that collectively painted murals, often overlooked in scholarship, are a valuable visual source to analyze contrasting ways of conceiving democracy from below and the appropriation of public spaces following the demise of two long-standing ultra-Catholic dictatorships. By acknowledging the multilayered power of images in murals and drawing on a rich body of research in visual studies, this contribution will address common people’s iconography in Iberian murals and the collective actions of those who sought to secure democratic rights. I will place particular emphasis not only on the visual representational framework of the relationship between the masses and individuals but also on the depiction of gender relations as an intrinsic part of the democratization process in both countries.

EnConVis: A Unified Framework for Ensemble Contour Visualization.

Ensemble simulation is a crucial method to handle potential uncertainty in modern simulation and has been widely applied in many disciplines. Many ensemble contour visualization methods have been introduced to facilitate ensemble data analysis. On the basis of deep exploration and summarization of existing techniques and domain requirements, we propose a unified framework of ensemble contour visualization, EnConVis (Ensemble Contour Visualization), which systematically combines state-of-the-art methods. We model ensemble contour visualization as a four-step pipeline consisting of four essential procedures: member filtering, point-wise modeling, uncertainty band extraction, and visual mapping. For each of the four essential procedures, we compare different methods they use, analyze their pros and cons, highlight research gaps, and attempt to fill them. Specifically, we add Kernel Density Estimation in the point-wise modeling procedure and multi-layer extraction in the uncertainty band extraction procedure. This step shows the ensemble datas details accurately and provides abstract levels. We also analyze existing methods from a global perspective. We investigate their mechanisms and compare their effects, on the basis of which, we offer selection guidelines for them. From the overall perspective of this framework, we find choices and combinations that have not been tried before, which can be well compensated by our method. Synthetic data and real-world data are leveraged to verify the efficacy of our method. Domain experts feedback suggests that our approach helps them better understand ensemble data analysis.

A deep generative framework for data-driven surrogate modeling and visualization of parameterized nonlinear dynamical systems

A deep generative framework for data-driven surrogate modeling and visualization of parameterized nonlinear dynamical systems

GRACE: a comprehensive web-based platform for integrative single-cell transcriptome analysis.

Large-scale single-cell RNA sequencing (scRNA-seq) has emerged as a robust method for dissecting cellular heterogeneity at single-cell resolution. However, to meet the increasingly high computational demands of non-programming experts, a user-friendly, scalable, and accessible online platform for analyzing scRNA-seq data is urgently needed. Here, we have developed a web-based platform GRACE (GRaphical Analyzing Cell Explorer) (http://grace.flowhub.com.cn or http://grace.jflab.ac.cn:28080) that enables online massive single-cell transcriptome analysis, improving interactivity and reproducibility using high-quality visualization frameworks. GRACE provides easy access to interactive visualization, customized parameters, and publication-quality graphs. Furthermore, it comprehensively integrates preprocessing, clustering, developmental trajectory inference, cell-cell communication, cell-type annotation, subcluster analysis, and pathway enrichment. In addition to the website platform, we also provide a Docker version that can be easily deployed on private servers. The source code for GRACE is freely available at (https://github.com/th00516/GRACE). Documentation and video tutorials are accessible from website homepage (http://grace.flowhub.com.cn). GRACE can analyze massive scRNA-seq data more flexibly and be accessible to the scientific community. This platform fulfills the major gap that exists between experimental (wet lab) and bioinformatic (dry lab) research.

Joint visualization of seasonal influenza serology and phylogeny to inform vaccine composition.

Seasonal influenza vaccines must be updated regularly to account for mutations that allow influenza viruses to escape our existing immunity. A successful vaccine should represent the genetic diversity of recently circulating viruses and induce antibodies that effectively prevent infection by those recent viruses. Thus, linking the genetic composition of circulating viruses and the serological experimental results measuring antibody efficacy is crucial to the vaccine design decision. Historically, genetic and serological data have been presented separately in the form of static visualizations of phylogenetic trees and tabular serological results to identify vaccine candidates. To simplify this decision-making process, we have created an interactive tool for visualizing serological data that has been integrated into Nextstrain's real-time phylogenetic visualization framework, Auspice. We show how the combined interactive visualizations may be used by decision makers to explore the relationships between complex data sets for both prospective vaccine virus selection and retrospectively exploring the performance of vaccine viruses.

A cyber physical production system framework for online monitoring, visualization and control by using cloud, fog, and edge computing technologies

ABSTRACT Cyber physical production system (CPPS) is an essential prerequisite to facilitate online monitoring, visualization, and control in a smart manufacturing system. The goal of this paper is to provide a proof of concept by proposing a CPPS framework, based on complementing cloud, fog and edge computing technologies, to demonstrate the real-time online monitoring, visualization, and control of a conventional 3D printer. The proposed methodology was successfully used for online detection and control of part defects, filament runouts, and smoke generation. The proposed methodology was also used to acquire, monitor, and visualize energy consumption, relative humidity, temperature, volatile organic compounds, particulate matters, and acceleration during 3D printing. The present work utilizes the computing technologies where they do not compete, but instead complement each other to enable intelligent capabilities and online sharing of resources which are scalable, reliable, and efficient. The proposed CPPS framework can be extended to provide low-cost solutions to micro, small and medium enterprises to convert their traditional critical equipment for real-time monitoring, visualization, control, and analytics.

Exploring view access for biophilic arctic architecture through immersive visualization of integrative lighting

This research addresses biophilic design opportunities in Arctic indoor environments through a comprehensive methodology that incorporates architectural features, extreme cold climate conditions, and integrative lighting attributes. The paper presents lighting and thermal conditions of higher latitudes and the potential challenges of biophilic window design in Arctic buildings. It discusses building standards and develops state-of-the-art techniques for analyzing quality views and lighting thresholds that support human health. It moreover proposes an innovative visualization framework to support decision-making in the design of indoor Arctic environments. A visualization process was developed and applied in environmental survey conducted through image-based and sensor-based techniques during an expedition in Iqaluktuuttiaq (Cambridge-Bay, 69°07′02″N 105°03′11″W), Nunavut in March 2020. The surveys combine the visualization of view access ratios in the 360° field of view (FOV), the intensity of melanopic/photopic luminance, and Equivalent Melanopic Lux (EML) through HDR imaging and radiometric measurements from several interior points at the Canadian High Arctic Research Station building. The paper also proposes 360° captures of surface temperatures as a compatible metric for the immersive visualization of physical ambiences in extreme cold climates. It compares view access ratios with human centric lighting potentials for 360° captures for each survey point. Results highlight the method's potential to support biophilic design in Arctic indoor spaces. It demonstrates how 360° FOV visualization of human responses to lighting conditions can provide a biophilic perspective of a space and its architectural properties.

Mobile augmented reality-based visualization framework for lifecycle O&M support of urban underground pipe networks

The structural deterioration of urban underground pipe networks is the accumulated combined effects of physical, environmental and operational factors during the whole lifecycle. Accurate on-site locating and mapping of pipes and their conditions has been a great challenge for construction, maintenance and emergency response in a dynamic manner. Existing approaches, mostly based on design-centric and static visualization, are unsuitable for implementation within the operation and maintenance (O&M) phase because they are performed in a completely virtual environment and are unable to interact with buried assets on site. An integrated hyperrealistic visualization framework based on mobile augmented reality (MAR) is developed to support various real-life project tasks for O&M support of urban underground pipe networks. The proposed approach consists of three parts: a cloud-based data management platform, a 3D visualization development platform, and a location-based registration method for MAR. A cloud–edge–mobile layer is proposed to manage, analyze, process and visualize multisource data, especially for dynamic data. Field experiments on a real-life urban gas pipe network were carried out to investigate the feasibility of the proposed framework to achieve informed O&M support. The results demonstrate that (1) a smartphone linked to a Lite RTK device allows for both accurate locating and easy operation; (2) both the 3D virtual digital model and the multisource data visually superimpose onto the realistic environment, enabling on-site human–machine interaction in a hyperrealistic manner; and (3) MAR-based immersive visualization contributes to avoiding excavation damage, improving O&M efficiency and assisting emergency responders to make timely and better–informed decisions based on multisource information fusion.