检索结果-内蒙古大学图书馆

Deep image captioning: A review of methods, trends and future challenges

NEUROCOMPUTING 2023年第1期546卷

作者： Xu, Liming Tang, Quan Lv, Jiancheng Zheng, Bochuan Zeng, Xianhua Li, Weisheng Sichuan Univ Coll Comp Sci Chengdu 610041 Peoples R China China West Normal Univ Sch Comp Sci Nanchong 637009 Peoples R China Chongqing Univ Posts & Telecommun Coll Comp Sci & Technol Chongqing 400065 Peoples R China

Image captioning, also called report generation in medical field, aims to describe visual content of images in human language, which requires to model semantic relationship between visual and textual elements and generate corresponding descriptions that conform to human language cognition. Image captioning is significant for promoting human-computer interaction in all fields and particularly, for computer-aided diagnosis in medical field. Currently, with the rapid development of deep learning technologies, image caption has attracted increasing attention of many researchers in artificial intelligence-related fields. To this end, this study attempts to provide readers with systematic and comprehensive research about different deep image captioning methods in natural and medical fields. We first introduce workflow of image captioning from perspective of simulating human process of describing images, including seeing, focusing and telling, which is respectively behavioralized into feature representation, visual encoding and language generation. Within it, we present common-used feature representation, visual encoding and language generation models. Then, we review datasets, evaluations and basic losses used in image cap-tioning, and summarize typical caption methods which are generally divided into that with or without using reinforcement learning. Besides, we describe advantages and disadvantages of existing methods, and conclusion and challenges are finally presented.(c) 2023 Published by Elsevier B.V.

关键词： Image caption Feature representation visual encoding Language generation Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Understanding Scholarly Neural Network System Diagrams Through Application of VisDNA 12th

Understanding Scholarly Neural Network System Diagrams Throu...

引用

12th International Conference on the Theory and Application of Diagrams (Diagrams)

作者： Marshall, Guy Clarke Jay, Caroline Freitas, Andre Univ Manchester Dept Comp Sci Manchester Lancs England Idiap Res Inst Martigny Switzerland

ISBN: (纸本)9783030860622;9783030860615

We utilise VisDNA as a tool for understanding neural network system architecture diagrams. Through examples from scholarly proceedings, we find that the application of the framework to this ecological and complex domain is effective for reflecting on these diagrams. We argue for additional vocabulary to describe semiotic variability and internal inconsistency or misuse of visual encoding principles in diagrams. Further, for application to system diagrams, we propose the addition of "Grouping by Object" as a new visual encoding principle, and "Emphasising" as a new visual encoding type.

关键词： Neural networks visual encoding Graphic language System diagrams

来源：评论

学校读者我要写书评

暂无评论

Estimating Color-Concept Associations from Image Statistics

引用

IEEE TRANSACTIONS ON visualIZATION AND COMPUTER GRAPHICS 2020年第1期26卷 1226-1235页

作者： Rathore, Ragini Leggon, Zachary Lessard, Laurent Schloss, Karen B. Univ Wisconsin Comp Sci Madison WI 53706 USA Univ Wisconsin WID Madison WI 53706 USA Univ Wisconsin Biol Madison WI 53706 USA Univ Wisconsin Elect & Comp Engn Madison WI 53706 USA Univ Wisconsin Psychol Madison WI 53706 USA

To interpret the meanings of colors in visualizations of categorical information, people must determine how distinct colors correspond to different concepts. This process is easier when assignments between colors and concepts in visualizations match people's expectations, making color palettes semantically interpretable. Efforts have been underway to optimize color palette design for semantic interpretablity, but this requires having good estimates of human color-concept associations. Obtaining these data from humans is costly, which motivates the need for automated methods. We developed and evaluated a new method for automatically estimating color-concept associations in a way that strongly correlates with human ratings. Building on prior studies using Google Images, our approach operates directly on Google Image search results without the need for humans in the loop. Specifically, we evaluated several methods for extracting raw pixel content of the images in order to best estimate color-concept associations obtained from human ratings. The most effective method extracted colors using a combination of cylindrical sectors and color categories in color space. We demonstrate that our approach can accurately estimate average human color-concept associations for different fruits using only a small set of images. The approach also generalizes moderately well to more complicated recycling-related concepts of objects that can appear in any color.

关键词： visual Reasoning visual Communication visual encoding Color Perception Color Cognition Color Categories

来源：评论

学校读者我要写书评

暂无评论

Decoding a Complex visualization in a Science Museum - An Empirical Study

引用

IEEE TRANSACTIONS ON visualIZATION AND COMPUTER GRAPHICS 2020年第1期26卷 472-481页

作者： Ma, Joyce Ma, Kwan-Liu Frazier, Jennifer Exploratorium San Francisco CA 94111 USA Univ Calif Davis Davis CA 95616 USA

This study describes a detailed analysis of museum visitors decoding process as they used a visualization designed to support exploration of a large, complex dataset. Quantitative and qualitative analyses revealed that it took, on average, 43 seconds for visitors to decode enough of the visualization to see patterns and relationships in the underlying data represented, and 54 seconds to arrive at their first correct data interpretation. Furthermore, visitors decoded throughout and not only upon initial use of the visualization. The study analyzed think-aloud data to identify issues visitors had mapping the visual representations to their intended referents, examine why they occurred, and consider if and how these decoding issues were resolved. The paper also describes how multiple visual encodings both helped and hindered decoding and concludes with implications on the design and adaptation of visualizations for informal science learning venues.

关键词： Museums informal science learning interactive exhibit public data visualization decoding visual encoding

来源：评论

学校读者我要写书评

暂无评论

OFF-transient alpha RGCs mediate looming triggered innate defensive response

引用

CURRENT BIOLOGY 2021年第11期31卷 2263-+页

作者： Wang, Fei Li, E. De, Lei Wu, Qiwen Zhang, Yifeng Chinese Acad Sci Inst Neurosci 320 Yueyang Rd Shanghai 200031 Peoples R China Univ Chinese Acad Sci 19 A Yuquan Rd Beijing 100049 Peoples R China Univ Calif San Francisco Dept Ophthalmol San Francisco CA 94143 USA Xiangtan Med & Hlth Vocat Coll 198 Shuangyong Middle Rd Xiangtan 411104 Hunan Peoples R China

Animals respond to visual threats, such as a looming object, with innate defensive behaviors. Here, we report that a specific type of retinal ganglion cell (RGC), the OFF-transient alpha RGC, is critical for the detection of looming objects. We identified Kcnip2 as its molecular marker. The activity of the Kcnip2-expressing RGCs encodes the size of the looming object. Ablation or suppression of these RGCs abolished or severely impaired the escape and freezing behaviors of mice in response to a looming object, while activation of their somas in the retina, or their axon terminals in the superior colliculus, triggered immediate escape behavior. Our results link the activity of a single type of RGC to visually triggered innate defensive behaviors and underscore that ethologically significant visual information is encoded by a labeled line strategy as early as in the retina.

关键词： looming visual threat visual encoding innate defensive response retinal ganglion cell αRGC superior colliculus labeled line

来源：评论

学校读者我要写书评

暂无评论

Mapping Color to Meaning in Colormap Data visualizations

引用

IEEE TRANSACTIONS ON visualIZATION AND COMPUTER GRAPHICS 2019年第1期25卷 810-819页

作者： Schloss, Karen B. Gramazio, Connor C. Silverman, Allison T. Parker, Madeline L. Wang, Audrey S. Univ Wisconsin Dept Psychol 1202 W Johnson St Madison WI 53706 USA Univ Wisconsin Wisconsin Inst Discovery Madison WI 53706 USA Brown Univ Dept Comp Sci Providence RI 02912 USA Brown Univ Sch Publ Hlth Providence RI 02912 USA CALTECH Dept Appl & Computat Math Pasadena CA 91125 USA

To interpret data visualizations, people must determine how visual features map onto concepts. For example, to interpret colormaps, people must determine how dimensions of color (e.g., lightness, hue) map onto quantities of a given measure (e.g., brain activity, correlation magnitude). This process is easier when the encoded mappings in the visualization match people's predictions of how visual features will map onto concepts, their inferred mappings. To harness this principle in visualization design, it is necessary to understand what factors determine people's inferred mappings. In this study, we investigated how inferred color-quantity mappings for colormap data visualizations were influenced by the background color. Prior literature presents seemingly conflicting accounts of how the background color affects inferred color-quantity mappings. The present results help resolve those conflicts, demonstrating that sometimes the background has an effect and sometimes it does not, depending on whether the colormap appears to vary in opacity. When there is no apparent variation in opacity, participants infer that darker colors map to larger quantities (dark-is-more bias). As apparent variation in opacity increases, participants become biased toward inferring that more opaque colors map to larger quantities (opaque-is-more bias). These biases work together on light backgrounds and conflict on dark backgrounds. Under such conflicts, the opaque-is-more bias can negate. or even supersede the dark-is-more bias. The results suggest that if a design goal is to produce colormaps that match people's inferred mappings and are robust to changes in background color, it is beneficial to use colormaps that will not appear to vary in opacity on any background color, and to encode larger quantities in darker colors.

关键词： visual Reasoning visual Communication Colormaps Color Perception visual encoding visual Design

来源：评论

学校读者我要写书评

暂无评论

Integrating visuospatial information across distinct experiences

引用

COGNITIVE PROCESSING 2019年第3期20卷 349-358页

作者： Pantelides, Stephanie N. Avraamides, Marios N. Univ Cyprus Dept Psychol Nicosia Cyprus Univ Cyprus Ctr Appl Neurosci POB 20537 Nicosia Cyprus

In two experiments, we examined whether the presence of stable visual information and the confluence of the viewpoints would cause participants to integrate in a single memory representation spatial locations they encoded at different points in time. Participants studied from the same or from different viewpoints two layouts of objects within a common visually cluttered room. Then, they carried out a series of pointing trials that involved objects from either the same or different layouts. Results showed that participants were faster for within- than between-layout judgments when they had studied the two layouts from different viewpoints but were equally fast across the two types of judgment after studying the layouts from the same viewpoint (Experiment 1). This finding suggests that they integrated locations into a single representation only when encoding the layouts from the same viewpoint. However, when participants' memory for the layout studied first was refreshed prior to testing (Experiment 2), no difference in response time was found, suggesting that they had integrated all locations in a single representation before the beginning of testing.

关键词： Integration Spatial cognition Spatial memory visual encoding Orientation

来源：评论

学校读者我要写书评

暂无评论

Assessing the Graphical Perception of Time and Speed on 2D+Time Trajectories

引用

IEEE TRANSACTIONS ON visualIZATION AND COMPUTER GRAPHICS 2018年第1期24卷 698-708页

作者： Perin, Charles Wun, Tiffany Pusch, Richard Carpendale, Sheelagh Univ London London England Univ Calgary Calgary AB Canada

We empirically evaluate the extent to which people perceive non-constant time and speed encoded on 2D paths. In our graphical perception study, we evaluate nine encodings from the literature for both straight and curved paths. visualizing time and speed information is a challenge when the x and y axes already encode other data dimensions, for example when plotting a trip on a map. This is particularly true in disciplines such as time-geography and movement analytics that often require visualizing spatio-temporal trajectories. A common approach is to use 2D+time trajectories, which are 2D paths for which time is an additional dimension. However, there are currently no guidelines regarding how to represent time and speed on such paths. Our study results provide InfoVis designers with clear guidance regarding which encodings to use and which ones to avoid;in particular, we suggest using color value to encode speed and segment length to encode time whenever possible.

关键词： Trajectory visualization visual encoding movement data graphical perception quantitative evaluation

来源：评论

学校读者我要写书评

暂无评论

A Framework for Analyzing and Designing Diagrams and Graphics 10th

A Framework for Analyzing and Designing Diagrams and Graphic...

引用

10th International Conference on the Theory and Application of Diagrams (Diagrams)

作者： Engelhardt, Yuri Richards, Clive Univ Twente Enschede Netherlands Birmingham City Univ Birmingham W Midlands England

ISBN: (纸本)9783319913766;9783319913759

A systematic method is presented that describes comprehensively the very broad design space of visualizations and the interrelationships between their constituents. The framework offered here includes three representational modes and fifteen visual encodings principles as the proposed universal building blocks of all types of diagrams and information graphics. The framework provides: 1. A vocabulary and a method for thoroughly analyzing the full spectrum of visual representations of information. 2. A mechanism for exploring previously unexploited combinations of visual encoding principles for representing information. 3. A potential tool for creating alternative representations for any given visualization or data set.

关键词： visual encoding Types of visual representations Design space Building blocks Graphic language Graphic relationships Nature of diagrams

来源：评论

学校读者我要写书评

暂无评论

Validation of SplitVectors encoding for Quantitative visualization of Large-Magnitude-Range Vector Fields

引用

IEEE TRANSACTIONS ON visualIZATION AND COMPUTER GRAPHICS 2017年第6期23卷 1691-1705页

作者： Zhao, Henan Bryant, Garnett W. Griffin, Wesley Terrill, Judith E. Chen, Jian Univ Maryland Dept Comp Sci & Elect Engn Baltimore MD 21250 USA NIST Gaithersburg MD 20899 USA

We designed and evaluated SplitVectors, a new vector field display approach to help scientists perform new discrimination tasks on large-magnitude-range scientific data shown in three-dimensional (3D) visualization environments. SplitVectors uses scientific notation to display vector magnitude, thus improving legibility. We present an empirical study comparing the SplitVectors approach with three other approaches - direct linear representation, logarithmic, and text display commonly used in scientific visualizations. Twenty participants performed three domain analysis tasks: reading numerical values (a discrimination task), finding the ratio between values (a discrimination task), and finding the larger of two vectors (a pattern detection task). Participants used both mono and stereo conditions. Our results suggest the following: (1) SplitVectors improve accuracy by about 10 times compared to linear mapping and by four times to logarithmic in discrimination tasks;(2) SplitVectors have no significant differences from the textual display approach, but reduce cluttering in the scene;(3) SplitVectors and textual display are less sensitive to data scale than linear and logarithmic approaches;(4) using logarithmic can be problematic as participants' confidence was as high as directly reading from the textual display, but their accuracy was poor;and (5) Stereoscopy improved performance, especially in more challenging discrimination tasks.

关键词： Vector field scientific visualization in virtual environments quantum physics visual encoding large-range data

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：