Categorizing art: Comparing humans and computers

The categorization of art (paintings, literature) into distinct styles such as Expressionism, or Surrealism has had a profound influence on how art is presented, marketed, analyzed, and historicized. Here, we present results from human and computational experiments with the goal of determining to which degree such categories can be explained by simple, low-level appearance information in the image. Following experimental methods from perceptual psychology on category formation, naive, non-expert participants were first asked to sort printouts of artworks from different art periods into categories. Converting these data into similarity data and running a multi-dimensional scaling (MDS) analysis, we found distinct categories which corresponded sometimes surprisingly well to canonical art periods. The result was cross-validated on two complementary sets of artworks for two different groups of participants showing the stability of art interpretation. The second focus of this paper was on determining how far computational algorithms would be able to capture human performance or would be able in general to separate different art categories. Using several state-of-the-art algorithms from computer vision, we found that whereas low-level appearance information can give some clues about category membership, human grouping strategies included also much higher-level concepts.     

New frontiers for information systems research: computer art as an information system

This paper proposes extending the frontiers of information systems (IS) research by seeing computer art as a kind of information system. It explores some research questions that arise if we take such a view. A new agenda for IS research is developed, based upon six themes in computer art and related work in previous IS research. The link between IS and computer art is bi-directional: our theories of information systems could inform the study of computer art, and computer art could inform our information systems theories, possibly leading to modification or even abandonment of them. Twelve research questions are suggested, and potential barriers to IS researchers studying computer art are also discussed. The paper offers possibilities, not solutions, and makes an important contribution by suggesting new opportunities for innovative research in IS and by signposting some of the topics which could be addressed.     

Browsing museum image collections on a multi-touch table

In this paper we present a multi-touch tabletop system for browsing image databases, conceived for museums and art gallery exhibitions. The system exploits an innovative image browsing paradigm and image retrieval functionalities to perform natural and intuitive user interaction: users can explore the image database by handling digital pictures by means of touch gestures or by the use of a predetermined set of physical objects; once one of these objects is placed on the table, it is automatically recognized and the associated function is triggered. The set of objects and the function associations can be dynamically configured. An innovative feature of our application is that users can interactively create and manipulate image clusters where images are grouped according to their pictorial similarity. This is achieved by placing one or more specific tangible objects on the table surface. The system has been evaluated on a collection of photos organized in groups according to the UNESCO picture categories. The usability tests, performed considering different user categories, show that users consider the application to be attractive and interesting.     

Pose-Invariant Face Alignment via CNN-Based Dense 3D Model Fitting

Pose-invariant face alignment is a very challenging problem in computer vision, which is used as a prerequisite for many facial analysis tasks, e.g., face recognition, expression recognition, and 3D face reconstruction. Recently, there have been a few attempts to tackle this problem, but still more research is needed to achieve higher accuracy. In this paper, we propose a face alignment method that aligns an image with arbitrary poses, by combining the powerful cascaded CNN regressors, 3D Morphable Model (3DMM), and mirrorability constraint. The core of our proposed method is a novel 3DMM fitting algorithm, where the camera projection matrix parameters and 3D shape parameters are estimated by a cascade of CNN-based regressors. Furthermore, we impose the mirrorability constraint during the CNN learning by employing a novel loss function inside the siamese network. The dense 3D shape enables us to design pose-invariant appearance features for effective CNN learning. Extensive experiments are conducted on the challenging large-pose face databases (AFLW and AFW), with comparison to the state of the art.     

基于创意序列学习的艺术风格学习与绘制系统

基于学习的智能图像风格绘制是当前多媒体特别是艺术风格化领域的一个热门研究课题.图像风格学习算法主要研究利用多媒体数据及机器学习方法对真实样本数据进行自动艺术绘制智能处理.目前主流方法主要针对艺术图像的静态样本进行学习.然而,由于静态数据类中包含的信息是平面化、局部化和非连续化的,难以保证风格化处理的全局一致性.本文旨在针对创意过程中的大规模复杂多媒体数据,从序列任务学习理论上提出一套智能艺术风格绘制的理论模型、设计方法以及优化方法三个层面.本文贡献主要体现在以下研究工作：（1）提出了一套面向数字美术的多媒体数据采集设备与软件系统,（2）利用IRL算法实现了艺术风格行为的模型化及其数字化保护方法,（3）提出基于PGPE的正则化策略学习方法以提高风格学习过程的稳定性.实验结果表明本文提出的方法行之有效地实现针对具体个性风格的照片水墨画艺术风格转化.基于序列化多媒体数据采集与分析,本文提出的面向移动互联网的自动艺术风格绘制辅助系统,不仅具有理论上的创新而且还具有实际上的巨大应用和价值.     

Superpixel Generation by Agglomerative Clustering With Quadratic Error Minimization

Superpixel segmentation is a popular image pre‐processing technique in many computer vision applications. In this paper, we present a novel superpixel generation algorithm by agglomerative clustering with quadratic error minimization. We use a quadratic error metric (QEM) to measure the difference of spatial compactness and colour homogeneity between superpixels. Based on the quadratic function, we propose a bottom‐up greedy clustering algorithm to obtain higher quality superpixel segmentation. There are two steps in our algorithm: merging and swapping. First, we calculate the merging cost of two superpixels and iteratively merge the pair with the minimum cost until the termination condition is satisfied. Then, we optimize the boundary of superpixels by swapping pixels according to their swapping cost to improve the compactness. Due to the quadratic nature of the energy function, each of these atomic operations has only O(1) time complexity. We compare the new method with other state‐of‐the‐art superpixel generation algorithms on two datasets, and our algorithm demonstrates superior performance.     

Artificial intelligence in Computer Aided Design

The combination of Artificial Intelligence (AI) and Computer Aided Design is a trend in computer aided techniques. In this paper, we discuss some AI techniques used in special CAD systems developed at Zhangjiang University. First, an expert system used in art pattern CAD is presented. Accessing the knowledge bases, the expert system reasons to produce various kinds of pseudo-3D art patterns which can be used for textile such as cotton, silk, and carpet. We shall focus on its knowledge representation and fuzzy reasoning mechanism. In the second part of the paper, a geometric modelling system for architectural CAD is described; and in particular, a method to construct the hierarchical geometry model of complex scene environment is introduced. The method is based on the clustering technique of Pattern Recognition. The geometric modelling system can build the hierarchical 3D geometric model automatically and the efficiency of the system is obviously improved with the method.     

SAR变化检测技术发展综述

合成孔径雷达(SAR)具有全天候、全天时的特点,是很好的变化检测信息源,研究SAR图像变化检测技术有着非常广阔的应用前景。系统介绍了SAR变化检测的整个流程--数据预处理、变化检测算法、检测后处理及变化检测算法评价。为了方便算法设计者,分门别类地阐述了变化检测算法的主要方法及核心决策规则,包括差值法、统计假设检验法、预测模型、相干模型;并对每种方法的特征进行了分析和比较。针对SAR技术的迅猛发展,我们对将来SAR变化检测技术的发展方向给予了预测。     

Doraemon as an artistic image processor

Doraemon is the name of a computer system thinking like a human which can behave phantasmagorically to create artistic expressions like metaphor. The conception of Doraemon's system is realized by Peirce's semiotic pragmatics, that is, the self-reflexive mechanism with meta-procedures which criticize and control its object-data-processing. Here we introduce (1) the pragmatically revised Frame and Actor and (2) the generative pragmatics derived from the pragmatically extended generative semantics and the CD theory to represent our creative Doraemon. The ideas and methods of Doraemon are expected to be effectively applied to creative education through art and to a new computer art.     

Real-time 3D fluid simulation digital art using BCI

We present a methodology about real-time 3D fluid simulation system using BCI (brain–computer interface). Spectators appreciate our work with their emotion. The advance of technology has exerted huge impacts on arts and as result a new genre called digital art has emerged. In digital art, interaction is an important element. Various types of interaction are being implemented in digital art using touch, hand gesture, sound, movement, etc., even brain wave. Usually, digital artists pursue interactions in their art work; the reason is that they want spectators’ deep appreciation and involvement. However, many of interactive digital art works fail to induce interaction as much as the artists expected. If visible interaction is weak, or interaction result is weak, spectators are not immersed much in the work. Nowadays, there are many interactive digital arts. In digital art, new type of interaction is needed to captivate spectators. So, we tried to implement BCI interaction digital art prototype (tentative title: Soul flow). Electroencephalography (EEG) is the recording of brain electrical activity along the scalp. We measured user’s EEG signal using MindSet (the MindSet senses EEG brainwave data to power the innovation of laboratory researchers and application developers like no other EEG device in the world. It delivers RAW signal, power frequency bands and NeuroSky eSense meters:attention, meditation. http://www.neurosky.com/products/mindset.aspx) to take user’s emotional state on appreciating digital art. We can get user’s emotional state in four categories (attention/inattention, meditation/uneasiness) and these emotional states interact with 3D fluid simulation system in real-time. In this art work prototype, each spectator experiences each different fluid (scale, speed and visual) because each spectator has different emotional state. Our art work prototype expressed the flow through water flow simulation. Flow means the mental state of operation in which a person performing an activity is fully immersed in a feeling of energized focus, full involvement, and enjoyment in the process of the activity. In essence, flow is characterized by complete absorption in what one does (The psychology of optimal experience 1990). In addition, it means stream in water flow. We pursued these concepts and expressed to 3D fluid simulation digital art using BCI. We used Unity3D and MindSet to implement our art work prototype. This study will expand the participation of the spectator in digital art. In addition, it will expand the possibilities of BCI in digital art interaction.     

Fully automatic person segmentation in unconstrained video using spatio-temporal conditional random fields

The segmentation of objects and people in particular is an important problem in computer vision. In this paper, we focus on automatically segmenting a person from challenging video sequences in which we place no constraint on camera viewpoint, camera motion or the movements of a person in the scene. Our approach uses the most confident predictions from a pose detector as a form of anchor or keyframe stick figure prediction which helps guide the segmentation of other more challenging frames in the video. Since even state of the art pose detectors are unreliable on many frames –especially given that we are interested in segmentations with no camera or motion constraints –only the poses or stick figure predictions for frames with the highest confidence in a localized temporal region anchor further processing. The stick figure predictions within confident keyframes are used to extract color, position and optical flow features. Multiple conditional random fields (CRFs) are used to process blocks of video in batches, using a two dimensional CRF for detailed keyframe segmentation as well as 3D CRFs for propagating segmentations to the entire sequence of frames belonging to batches. Location information derived from the pose is also used to refine the results. Importantly, no hand labeled training data is required by our method. We discuss the use of a continuity method that reuses learnt parameters between batches of frames and show how pose predictions can also be improved by our model. We provide an extensive evaluation of our approach, comparing it with a variety of alternative grab cut based methods and a prior state of the art method. We also release our evaluation data to the community to facilitate further experiments. We find that our approach yields state of the art qualitative and quantitative performance compared to prior work and more heuristic alternative approaches.     

An asymmetric and optimized encryption method to protect the confidentiality of 3D mesh model

3D models are widely used in computer graphics, design and manufacture engineering, art animation and entertainment. With the universal of acquisition equipment and sensors, a huge number of 3D models are generated, which are becoming the major source of engineering data. How to preserve the privacy of the 3D models is a challenge issue. In this paper, an asymmetric and optimized encryption method is presented to protect the 3D mesh models. Firstly, we propose an asymmetric encryption method for 3D mesh models to overcome the drawbacks of traditional symmetric encryption. The primary benefit is that our approach can enhance the security of the key. Secondly, we extend the typically asymmetric encryption algorithm from integer domain to float domain. In our method, we present a normalization function to map the float DC (Discrete Cosine) coefficients to integer domain. Thirdly, considering that the shape error and encryption/decryption computation cost are contradictory in the normalization mapping, we formulate the contradiction as a multi-objective optimization problem. And then, we propose a multi-objective solution to find an optimized mapping range for encryption/decryption efficiently. Furthermore, benefiting from the proposed asymmetric encryption framework, we continue to put forward a method to check the integrity of the encrypted 3D mesh model, in which the digest is encrypted twice to generate digital signature more safely. The proposed method has been tested on 3D mesh models from Stanford university and other sources to demonstrate the effect of the proposed encryption method and optimization mechanism.     

一种基于改进ART-2的入侵检测方法

指出了直接把标准ART-2网络应用于入侵检测时存在的两个问胚：对基本相似、仅有个别分量差别较大的向量不能正确分类；输入向量特征丢失。并根据入侵检测的特定应用，相应地提出在首先对输入向量进行规范化处理．然后用新引入的一种具有更严格测试准则ART-2网络对其进行处理的方法，以期提高入侵检测系统的检测率和误检率。     

Hybrid system of ART and RBF neural networks for online clustering

An online clustering task is considered for machine state monitoring purpose. In the previous authors’ researches a classical ART-2 network was tested for online classification of operational states in the context of a wind turbine monitoring. Some drawbacks, however, were found when a data stream size had been increased. This case is investigated in this paper. Classical ART-2 network can cluster data incorrectly when data points are compared by using Euclidean distance. Furthermore, ART-2 network can lose accuracy when data stream is processed for long time. The way of improving the ART-2 network is considered and two main steps of that are taken. At first, the stereographic projection is implemented. At the second step, a new type of hybrid neural system which consists of ART-2 and RBF networks with data processed by using the stereographic projection is introduced. Tests contained elementary scenarios for low-dimensional cases as well as higher dimensional real data from wind turbine monitoring. All the tests implied that an efficient system for online clustering had been found.     

ART-2神经网络的改进及建模实现

指出了传统的ART-2神经网络对渐变过程不敏感的局限性,提出了一种新的改进算法。并对ART-2网络进行建模,通过与其它建模方法的对比,详尽讨论了ART-2的建模方法及特点。最后通过应用改进算法解决了原先模型中的“模式漂移”现象,使模型性能得到了明显的改善。     

基于实时技术的大气颗粒物在线分析系统

针对环境监测中,难以实时在线处理海量颗粒物数据的问题,提出了一种基于实时技术的大气颗粒物在线分析系统,实现了颗粒物统计、浓度变化、来源解析等功能.该系统利用实时数据库来实时采集、存储海量大气数据,解决了环境监测中数据的海量问题;同时,引入自适应共振神经网络算法和逻辑回归模型进行数据分析,成功降低数据规模,提升数据分析速度.实践表明,该在线分析系统能在合理时间内得到准确的分析结果,具有重要的实际意义.     

基于图像语义分割的真实毛笔笔触实时生成技术

书法在文化传承中占据重要地位,书法书写笔迹的生成也一直是计算机图形学的研究重点和难点.现存基于模型和经验的方法,由于建模难度大,大都将笔触表述为简单的几何图形并且缺少变化,难以真实还原毛笔书写的笔触和笔迹.使得现存书法笔迹生成软件仅仅用于娱乐,而难以上升到数字化书法教育层面.文中从计算机视觉的角度出发,通过4个相机获取毛笔的实时书写图像;针对Deeplabv3+语义分割算法无法有效地分割小尺寸类别的缺点进行优化,使用优化的Deeplabv3+算法提取图像中毛笔笔头等关键信息,并通过Hough变换和PnP位姿估计算法计算笔杆相对位姿;基于位姿信息矫正和融合各相机笔触图像,提出一种未知区域估计方法估计相机无法拍摄到的笔触区域.按照不同条件提取400多幅书写图像作为数据集并进行实验结果表明,优化后的Deeplabv3+算法平均交并比(mean intersection-over-union,mIOU)达到0.849,与优化前相比提升了0.117;在小尺寸类别上交并比(intersection-over-union,IOU)达到0.59,提升了0.473.在保证实时性的前提下,最终生成的笔触与传统基于模型和经验的方法相比,可以更加真实地还原书写时的笔触,并避免对毛笔进行复杂的建模,为笔迹生成研究提供一种新的思路.     

Computers and Art

The use of computers in art is not new: examples are given of past and present applications. Three categories are mentioned: computer as tool, as origin of concepts and metaphors, and as controller. The use of the computer in art gives rise to the idea of computing as representation- rather than just information-processing. Thus the computer can be said to 'belong' to artists and designers, and can be seen as a qualitative device. It is argued that computing is 'really' an art or design activity, and that this has implications for art, design and computing.     

LSMAT Least Squares Medial Axis Transform

The medial axis transform has applications in numerous fields including visualization, computer graphics, and computer vision. Unfortunately, traditional medial axis transformations are usually brittle in the presence of outliers, perturbations and/or noise along the boundary of objects. To overcome this limitation, we introduce a new formulation of the medial axis transform which is naturally robust in the presence of these artefacts. Unlike previous work which has approached the medial axis from a computational geometry angle, we consider it from a numerical optimization perspective. In this work, we follow the definition of the medial axis transform as ‘the set of maximally inscribed spheres’. We show how this definition can be formulated as a least squares relaxation where the transform is obtained by minimizing a continuous optimization problem. The proposed approach is inherently parallelizable by performing independent optimization of each sphere using Gauss–Newton, and its least‐squares form allows it to be significantly more robust compared to traditional computational geometry approaches. Extensive experiments on 2D and 3D objects demonstrate that our method provides superior results to the state of the art on both synthetic and real‐data.     

Robust 3D object registration without explicit correspondence using geometric integration

3D vision-guided manipulation of components is a key problem of industrial machine vision. In this paper, we focus on the localization and pose estimation of known industrial objects from 3D measurements delivered by a scanning sensor. Since local information extracted from these measurements is unreliable due to noise, spatially unstructured measurements and missing detections, we present a novel objective function for robust registration without using correspondence information, based on the likelihood of model points. Furthermore, by extending Runge–Kutta-type integration directly to the group of Euclidean transformation, we infer object pose by computing the gradient flow directly on the related manifold. Comparison of our approach to existing state of the art methods shows that our method is more robust against poor initializations while having comparable run-time performance.