Selfie retoucher: subject-oriented self-portrait enhancement

Sharing self-portraits starts trending nowadays with the boom of social networks and the rise of smartphones. However, limited by the hardware capabilities, self-portraits taken by the front cameras of portable media devices usually face quality problems such as an incomplete field of view and poor lighting style. In our paper, we introduce a selfie retoucher which enhances a self-portrait with the help of N supporting photos that share the same scene and similar shooting time. With the extra information brought by the supporting photos, a lager field of view and a better lighting style can be achieved. To accomplish this, we propose a novel subject-oriented self-portrait enhancement method with a cascaded illumination unification and photos registration framework. Based on the correspondences extracted from the input 1+N photos, our method estimates and updates the illumination and registration coefficients in a cascaded manner. Moreover, a subject-oriented enhancement algorithm is proposed to enhance the face of the photographer in the self-portrait. We adopt a face-specific illumination correction process over the self-portrait to further improve the visual quality of the subject. After the enhancement, we globally fuse the aligned photos by a Markov Random Field based optimization method. During the fusion, a body map is additionally derived from the subject for guidance. Experimental results demonstrate that the proposed method achieves high-quality results in this novel application scenario.

     

Model-driven development of asynchronous message-driven architectures with AsyncAPI

In the Internet-of-Things (IoT) vision, everyday objects evolve into cyber-physical systems. The massive use and deployment of these systems has given place to the Industry 4.0 or Industrial IoT (IIoT). Due to its scalability requirements, IIoT architectures are typically distributed and asynchronous. In this scenario, one of the most widely used paradigms is publish/subscribe, where messages are sent and received based on a set of categories or topics. However, these architectures face interoperability challenges. Consistency in message categories and structure is the key to avoid potential losses of information. Ensuring this consistency requires complex data processing logic both on the publisher and the subscriber sides. In this paper, we present our proposal relying on AsyncAPI to automate the design and implementation of these asynchronous architectures using model-driven techniques for the generation of (part of) message-driven infrastructures. Our proposal offers two different ways of designing the architectures: either graphically, by modeling and annotating the messages that are sent among the different IoT devices, or textually, by implementing an editor compliant with the AsyncAPI specification. We have evaluated our proposal by conducting a set of experiments with 25 subjects with different expertise and background. The experiments show that one-third of the subjects were able to design and implement a working architecture in less than an hour without previous knowledge of our proposal, and an additional one-third estimated that they would only need less than two hours in total.

     

Sparse sample self-representation for subspace clustering

This paper proposes a new subspace clustering method based on sparse sample self-representation (SSR). The proposed method considers SSR to solve the problem that affinity matrix does not strictly follow the structure of subspace, and also utilizes sparse constraint to ensure the robustness to noise and outliers in subspace clustering. Specifically, we propose to first construct a self-representation matrix for all samples and combine an l ₁-norm regularizer with an l _2,1-norm regularizer to guarantee that each sample can be represented as a sparse linear combination of its related samples. Then, we conduct the resulting matrix to build an affinity matrix. Finally, we apply spectral clustering on the affinity matrix to conduct clustering. In order to validate the effectiveness of the proposed method, we conducted experiments on UCI datasets, and the experimental results showed that our proposed method reduced the minimal clustering error, outperforming the state-of-the-art methods.

     

Avoiding unnecessary information loss: correct and efficient model synchronization based on triple graph grammars

Model synchronization, i.e., the task of restoring consistency between two interrelated models after a model change, is a challenging task. Triple graph grammars (TGGs) specify model consistency by means of rules that describe how to create consistent pairs of models. These rules can be used to automatically derive further rules, which describe how to propagate changes from one model to the other or how to change one model in such a way that propagation is guaranteed to be possible. Restricting model synchronization to these derived rules, however, may lead to unnecessary deletion and recreation of model elements during change propagation. This is inefficient and may cause unnecessary information loss, i.e., when deleted elements contain information that is not represented in the second model, this information cannot be recovered easily. Short-cut rules have recently been developed to avoid unnecessary information loss by reusing existing model elements. In this paper, we show how to automatically derive (short-cut) repair rules from short-cut rules to propagate changes such that information loss is avoided and model synchronization is accelerated. The key ingredients of our rule-based model synchronization process are these repair rules and an incremental pattern matcher informing about suitable applications of them. We prove the termination and the correctness of this synchronization process and discuss its completeness. As a proof of concept, we have implemented this synchronization process in eMoflon, a state-of-the-art model transformation tool with inherent support of bidirectionality. Our evaluation shows that repair processes based on (short-cut) repair rules have considerably decreased information loss and improved performance compared to former model synchronization processes based on TGGs.

     

l1优化在网格去噪中的应用

目的 网格去噪是计算机图形学中的经典问题,而如何在去除噪声的同时保持网格的特征结构是这一研究方向所面临的最大挑战。方法 提出一种具有稀疏性的全局网格去噪方法,该方法源于信号处理理论中稀疏表示的基本思想,通过优化全局能量函数来去除网格模型的噪声,同时能够保持网格模型的特征结构。该方法共分为两个步骤,第1步为网格面法向量的滤波,首先建立全局优化模型,对噪声网格的面法向量进行滤波优化,其中引入l₁范数来保证解的稀疏性,使得优化后新的面法向量能够保持网格的特征结构;第2步为网格曲面的重建,根据第1步得到的新的面法向量,按照面法向量的定义,建立最小二乘意义下的网格顶点的重建模型,求解得到新的网格曲面。结果 由于该模型是全局方法,避免了现有滤波方法可能出现的不收敛等问题,能够取得比较满意的去噪效果。结论 大量实验结果表明,本文方法在去除噪声的同时,能较好地保持网格的特征结构,尤其对于CAD模型有很好的实验效果。     

Attacks on state-of-the-art face recognition using attentional adversarial attack generative network

With the broad use of face recognition, its weakness gradually emerges that it is able to be attacked. Therefore, it is very important to study how face recognition networks are subject to attacks. Generating adversarial examples is an effective attack method, which misleads the face recognition system through obfuscation attack (rejecting a genuine subject) or impersonation attack (matching to an impostor). In this paper, we introduce a novel GAN, Attentional Adversarial Attack Generative Network (A³GN), to generate adversarial examples that mislead the network to identify someone as the target person not misclassify inconspicuously. For capturing the geometric and context information of the target person, this work adds a conditional variational autoencoder and attention modules to learn the instance-level correspondences between faces. Unlike traditional two-player GAN, this work introduces a face recognition network as the third player to participate in the competition between generator and discriminator which allows the attacker to impersonate the target person better. The generated faces which are hard to arouse the notice of onlookers can evade recognition by state-of-the-art networks and most of them are recognized as the target person.

     

Parametric Generation of Facial Expressions Based on FACS

This paper presents a parametric performance‐based model for facial animation that was inspired by Facial Action Coding System (FACS) developed by P. Ekman and F. W. Friesen. The FACS consists of 44 Action Units (AUs) corresponding to visual changes on the face. Additionally, predefined co‐occurrence rules describe how different AUs influence each other. In our model, each facial animation parameter corresponds to one of the AUs as defined in FACS. Implementation of the model is completed with methods for accumulating displacement from separate AUs together, and fuzzy‐logical adaptation of co‐occurrence rules from the FACS. We also describe the method for adapting our model to a specific person.     

Prediction of Individual Non-Linear Aging Trajectories of Faces

Represented in a Morphable Model, 3D faces follow curved trajectories in face space as they age. We present a novel algorithm that computes the individual aging trajectories for given faces, based on a non-linear function that assigns an age to each face vector. This function is learned from a database of 3D scans of teenagers and adults using support vector regression. To apply the aging prediction to images of faces, we reconstruct a 3D model from the input image, apply the aging transformation on both shape and texture, and then render the face back into the same image or into images of other individuals at the appropriate ages, for example images of older children. Among other applications, our system can help to find missing children.     

What's in a Face? Metric Learning for Face Characterization

We present a method for determining which facial parts (mouth, nose, etc.) best characterize an individual, given a set of that individual's portraits. We introduce a novel distinctiveness analysis of a set of portraits, which leverages the deep features extracted by a pre‐trained face recognition CNN and a hair segmentation FCN, in the context of a weakly supervised metric learning scheme. Our analysis enables the generation of a polarized class activation map (PCAM) for an individual's portrait via a transformation that localizes and amplifies the discriminative regions of the deep feature maps extracted by the aforementioned networks. A user study that we conducted shows that there is a surprisingly good agreement between the face parts that users indicate as characteristic and the face parts automatically selected by our method. We demonstrate a few applications of our method, including determining the most and the least representative portraits among a set of portraits of an individual, and the creation of facial hybrids: portraits that combine the characteristic recognizable facial features of two individuals. Our face characterization analysis is also effective for ranking portraits in order to find an individual's look‐alikes (Doppelgängers).     

Completeness and Counter-Example Generations of a Basic Protocol Logic: (Extended Abstract)

We give an axiomatic system in first-order predicate logic with equality for proving security protocols correct. Our axioms and inference rules derive the basic inference rules, which are explicitly or implicitly used in the literature of protocol logics, hence we call our axiomatic system Basic Protocol Logic (or BPL, for short). We give a formal semantics for BPL, and show the completeness theorem such that for any given query (which represents a correctness property) the query is provable iff it is true for any model. Moreover, as a corollary of our completeness proof, the decidability of provability in BPL holds for any given query. In our formal semantics we consider a “trace” any kind of sequence of primitive actions, counter-models (which are generated from an unprovable query) cannot be immediately regarded as realizable traces (i.e., attacked processes on the protocol in question). However, with the aid of Comon-Treinen's algorithm for the intruder deduction problem, we can determine whether there exists a realizable trace among formal counter-models, if any, generated by the proof-search method (used in our completeness proof). We also demonstrate that our method is useful for both proof construction and flaw analysis by using a simple example.     

Provably correct and complete transaction rules for updating 3D city models

The shapes of our cities change very frequently. These changes have to be reflected in data sets representing urban objects. However, it must be assured that frequent updates do not affect geometric-topological consistency. This important aspect of spatial data quality guarantees essential assumptions on which users and applications of 3D city models rely: viz. that objects do not intersect, overlap or penetrate mutually, or completely cover one another. This raises the question how to guarantee that geometric-topological consistency is preserved when data sets are updated. Hence, there is a certain risk that plans and decisions which are based on these data sets are erroneous and that the tremendous efforts spent for data acquisition and updates become vain. In this paper, we solve this problem by presenting efficient transaction rules for updating 3D city models. These rules guarantee that geometric-topological consistency is preserved (Safety) and allow for the generation of arbitrary consistent 3D city models (Completeness). Safety as well as completeness is proven with mathematical rigor, guaranteeing the reliability of our method. Our method is applicable to 3D city models, which define—besides the terrain surface—complex spatial objects like buildings with rooms and storeys as interior structures, as well as bridges and tunnels. Those objects are represented as aggregations of solids, and their surfaces are complex from a topology point of view. 3D GIS models like CityGML, which are widely used to represent cities, provide the means to define semantics, geometry and topology, but do not address the problem of maintaining consistency. Hence, our approach complements CityGML.     

Efficient Graph-Based Algorithms for Discovering and Maintaining Association Rules in Large Databases

In this paper, we study the issues of mining and maintaining association rules in a large database of customer transactions. The problem of mining association rules can be mapped into the problems of finding large itemsets which are sets of items brought together in a sufficient number of transactions. We revise a graph-based algorithm to further speed up the process of itemset generation. In addition, we extend our revised algorithm to maintain discovered association rules when incremental or decremental updates are made to the databases. Experimental results show the efficiency of our algorithms. The revised algorithm is a significant improvement over the original one on mining association rules. The algorithms for maintaining association rules are more efficient than re-running the mining algorithms for the whole updated database and outperform previously proposed algorithms that need multiple passes over the database. Received 4 August 1999 / Revised 18 March 2000 / Accepted in revised form 18 October 2000     

An Efficient Framework for Multiple Subgraph Pattern Matching Models

With the popularity of storing large data graph in cloud, the emergence of subgraph pattern matching on a remote cloud has been inspired. Typically, subgraph pattern matching is defined in terms of subgraph isomorphism, which is an NP-complete problem and sometimes too strict to find useful matches in certain applications. And how to protect the privacy of data graphs in subgraph pattern matching without undermining matching results is an important concern. Thus, we propose a novel framework to achieve the privacy-preserving subgraph pattern matching in cloud. In order to protect the structural privacy in data graphs, we firstly develop a k-automorphism model based method. Additionally, we use a cost-model based label generalization method to protect label privacy in both data graphs and pattern graphs. During the generation of the k-automorphic graph, a large number of noise edges or vertices might be introduced to the original data graph. Thus, we use the outsourced graph, which is only a subset of a k-automorphic graph, to answer the subgraph pattern matching. The efficiency of the pattern matching process can be greatly improved in this way. Extensive experiments on real-world datasets demonstrate the high efficiency of our framework.

     

Implicit relative attribute enabled cross-modality hashing for face image-video retrieval

Face image-video retrieval refers to retrieving videos of a specific person with image query or searching face images of one person by using a video clip query. It has attracted much attention for broad applications like suspect tracking and identifying. This paper proposes a novel implicit relative attribute enabled cross-modality hashing (IRAH) method for large-scale face image-video retrieval. To cope with large-scale data, the proposed IRAH method facilitates fast cross-modality retrieval through embedding two entirely heterogeneous spaces, i.e., face images in Euclidean space and face videos on a Riemannian manifold, into a unified compact Hamming space. In order to resolve the semantic gap, IRAH maps the original low-level kernelized features to discriminative high-level implicit relative attributes. Therefore, the retrieval accuracy can be improved by leveraging both the label information across different modalities and the semantic structure obtained from the implicit relative attributes in each modality. To evaluate the proposed method, we conduct extensive experiments on two publicly available databases, i.e., the Big Bang Theory (BBT) and Buffy the Vampire Slayer (BVS). The experimental results demonstrate the superiority of the proposed method over different state-of-the-art cross-modality hashing methods. The performance gains are especially significant in the case that the hash code length is 8 bits, up to 12% improvements over the second best method among tested methods.

     

SELF-ORGANIZATION OF LIVING SYSTEMS: A FORMAL MODEL OF AUTOPOIESIS

ABSTRACT

A formalization, computerization and extension of the original Varela-Maturana-Uribe model of autopoiesis is presented. Autopoietic systems are driven by sets of simple “rules” which guide the behavior of components in a given milieu. These rules are capable of producing systemic structures that are far more complex than we could ever achieve by a direct arrangement of components, i.e., by a method of systems analysis and design. The study of autopoietic systems indicates that the traditional emphasis on internal qualities of system's components has been misplaced. It is the organization of components, rather than the components themselves (or their structural manifestations), that provides the necessary and sufficient conditions of autopoiesis and thus of life itself. The dynamic autonomy of autopoietic systems contrasts significantly with the non-autonomous, allopoietic mechanistic systems.     

结合形变模型与图像修复的人脸姿态矫正

目的 人脸姿态偏转是影响人脸识别准确率的一个重要因素,本文利用3维人脸重建中常用的3维形变模型以及深度卷积神经网络,提出一种用于多姿态人脸识别的人脸姿态矫正算法,在一定程度上提高了大姿态下人脸识别的准确率。方法 对传统的3维形变模型拟合方法进行改进,利用人脸形状参数和表情参数对3维形变模型进行建模,针对面部不同区域的关键点赋予不同的权值,加权拟合3维形变模型,使得具有不同姿态和面部表情的人脸图像拟合效果更好。然后,对3维人脸模型进行姿态矫正并利用深度学习对人脸图像进行修复,修复不规则的人脸空洞区域,并使用最新的局部卷积技术同时在新的数据集上重新训练卷积神经网络,使得网络参数达到最优。结果 在LFW（labeled faces in the wild）人脸数据库和StirlingESRC（Economic Social Research Council）3维人脸数据库上,将本文算法与其他方法进行比较,实验结果表明,本文算法的人脸识别精度有一定程度的提高。在LFW数据库上,通过对具有任意姿态的人脸图像进行姿态矫正和修复后,本文方法达到了96.57%的人脸识别精确度。在StirlingESRC数据库上,本文方法在人脸姿态为±22°的情况下,人脸识别准确率分别提高5.195%和2.265%;在人脸姿态为±45°情况下,人脸识别准确率分别提高5.875%和11.095%;平均人脸识别率分别提高5.53%和7.13%。对比实验结果表明,本文提出的人脸姿态矫正算法有效提高了人脸识别的准确率。结论 本文提出的人脸姿态矫正算法,综合了3维形变模型和深度学习模型的优点,在各个人脸姿态角度下,均能使人脸识别准确率在一定程度上有所提高。     

CONFIDENCE LIMITS ON ROUGH SET RULES GENERATED FROM A LIMITED NUMBER OF EXAMPLES

Abstract

The rough sets method is used for extracting both certain and possible rules from data. This paper shows that, in reality, there are no certain rules. Probability theory is used to determine the best distribution to use when evaluating the sirength of rules. A method of determining the confidence limits for rules is presented, and this is used to determine what rule to follow when conflicts occur. Finally, a way to apply these results to situations where the cost of wrong decisions is different from the rewards for correct decisions is discussed.     

A context-aware recommender method based on text and opinion mining

A recommender system is an information filtering technology that can be used to recommend items that may be of interest to users. Additionally, there are the context-aware recommender systems that consider contextual information to generate the recommendations. Reviews can provide relevant information that can be used by recommender systems, including contextual and opinion information. In a previous work, we proposed a context-aware recommendation method based on text mining (CARM-TM). The method includes two techniques to extract context from reviews: CIET.5_embed, a technique based on word embeddings; and RulesContext, a technique based on association rules. In this work, we have extended our previous method by including CEOM, a new technique which extracts context by using aspect-based opinions. We call our extension of CARM-TOM (context-aware recommendation method based on text and opinion mining). To generate recommendations, our method makes use of the CAMF algorithm, a context-aware recommender based on matrix factorization. To evaluate CARM-TOM, we ran an extensive set of experiments in a dataset about restaurants, comparing CARM-TOM against the MF algorithm, an uncontextual recommender system based on matrix factorization; and against a context extraction method proposed in literature. The empirical results strongly indicate that our method is able to improve a context-aware recommender system.     

Abstract meaning representation for legal documents: an empirical research on a human-annotated dataset

Natural language processing techniques contribute more and more in analyzing legal documents recently, which supports the implementation of laws and rules using computers. Previous approaches in representing a legal sentence often based on logical patterns that illustrate the relations between concepts in the sentence, often consist of multiple words. Those representations cause the lack of semantic information at the word level. In our work, we aim to tackle such shortcomings by representing legal texts in the form of abstract meaning representation (AMR), a graph-based semantic representation that gains lots of polarity in NLP community recently. We present our study in AMR Parsing (producing AMR from natural language) and AMR-to-text Generation (producing natural language from AMR) specifically for legal domain. We also introduce JCivilCode, a human-annotated legal AMR dataset which was created and verified by a group of linguistic and legal experts. We conduct an empirical evaluation of various approaches in parsing and generating AMR on our own dataset and show the current challenges. Based on our observation, we propose our domain adaptation method applying in the training phase and decoding phase of a neural AMR-to-text generation model. Our method improves the quality of text generated from AMR graph compared to the baseline model. (This work is extended from our two previous papers: “An Empirical Evaluation of AMR Parsing for Legal Documents”, published in the Twelfth International Workshop on Juris-informatics (JURISIN) 2018; and “Legal Text Generation from Abstract Meaning Representation”, published in the 32nd International Conference on Legal Knowledge and Information Systems (JURIX) 2019.).

     

Face anti-spoofing with generated near-infrared images

Reflection differences between live faces and spoof faces under near-infrared spectrum make near-infrared image based methods obtain superior performance for face anti-spoofing. However, for conventional face recognition systems, near-infrared image based methods need additional near-infrared equipment to capture the input near-infrared images. In this paper, we propose a novel face anti-spoofing method which exploits the clues in both visible light (VIS) images and near-infrared (NIR) images without utilizing any near-infrared equipment during testing. Specifically, we first propose a novel multiple categories image translation generative adversarial network (MCT-GAN) which generates corresponding NIR images for VIS live and spoof face images. Then we utilize convolution neural network to learn fusing features from both VIS images and corresponding generated NIR images for the goal of live and spoof face classification. Qualitative and quantitative experiments demonstrate that our method obtains excellent results compared to the state-of-the-art methods.