scispace - formally typeset
Search or ask a question

Showing papers on "Signature recognition published in 2010"


Journal ArticleDOI
01 May 2010
TL;DR: An approach is proposed that is able to guarantee security and renewability to biometric templates, which can be applied to any biometrics whose template can be represented by a set of sequences, in order to generate multiple transformed versions of the template.
Abstract: Recent years have seen the rapid spread of biometric technologies for automatic people recognition. However, security and privacy issues still represent the main obstacles for the deployment of biometric-based authentication systems. In this paper, we propose an approach, which we refer to as BioConvolving, that is able to guarantee security and renewability to biometric templates. Specifically, we introduce a set of noninvertible transformations, which can be applied to any biometrics whose template can be represented by a set of sequences, in order to generate multiple transformed versions of the template. Once the transformation is performed, retrieving the original data from the transformed template is computationally as hard as random guessing. As a proof of concept, the proposed approach is applied to an on-line signature recognition system, where a hidden Markov model-based matching strategy is employed. The performance of a protected on-line signature recognition system employing the proposed BioConvolving approach is evaluated, both in terms of authentication rates and renewability capacity, using the MCYT signature database. The reported extensive set of experiments shows that protected and renewable biometric templates can be properly generated and used for recognition, at the expense of a slight degradation in authentication performance.

157 citations


Journal ArticleDOI
TL;DR: An on-line signature authentication system based on an ensemble of local, regional, and global matchers is presented and a template protection scheme employing the BioHashing and the BioConvolving approaches, two well known template protection techniques for biometric recognition, is discussed.
Abstract: In this work an on-line signature authentication system based on an ensemble of local, regional, and global matchers is presented. Specifically, the following matching approaches are taken into account: the fusion of two local methods employing Dynamic Time Warping, a Hidden Markov Model based approach where each signature is described by means of its regional properties, and a Linear Programming Descriptor classifier trained by global features. Moreover, a template protection scheme employing the BioHashing and the BioConvolving approaches, two well known template protection techniques for biometric recognition, is discussed. The reported experimental results, evaluated on the public MCYT signature database, show that our best ensemble obtains an impressive Equal Error Rate of 3%, when only five genuine signatures are acquired for each user during enrollment. Moreover, when the proposed protected system is taken into account, the Equal Error Rate achieved in the worst case scenario, that is,when an ''impostor'' is able to steal the hash keys, is equal to 4.51%, whereas an Equal Error Rate ~0 can be obtained when nobody steals the hash keys.

77 citations


Journal ArticleDOI
TL;DR: The target of research is to present online handwritten signature verification system based on discrete wavelet transform (DWT) features extraction and feed forward back propagation error neural network recognition.

68 citations


Journal ArticleDOI
TL;DR: The reported results show that, when using the proposed cryptosystem, protected and renewable signature templates can be properly generated and used for recognition purposes.
Abstract: In this paper we present a novel on-line signature based biometric recognition system, where cryptographic techniques are employed to provide protection and cancelability to function based signature templates. The performances of the proposed protected on-line signature recognition system are evaluated on the public MCYT signature database, and compared with the performances achievable using unprotected approaches, as well as other signature template protection approaches. The reported results show that, when using the proposed cryptosystem, protected and renewable signature templates can be properly generated and used for recognition purposes.

60 citations


Dissertation
01 Jan 2010
TL;DR: A log-linear modeling framework is established in the context of discriminative training criteria, with examples from continuous speech recognition, part-of-speech tagging, and handwriting recognition, and the focus will be on the theoretical and experimental comparison of different training algorithms.
Abstract: Conventional speech recognition systems are based on Gaussian hidden Markov models (HMMs). Discriminative techniques such as log-linear modeling have been investigated in speech recognition only recently. This thesis establishes a log-linear modeling framework in the context of discriminative training criteria, with examples from continuous speech recognition, part-of-speech tagging, and handwriting recognition. The focus will be on the theoretical and experimental comparison of different training algorithms. Equivalence relations for Gaussian and log-linear models in speech recognition are derived. It is shown how to incorporate a margin term into conventional discriminative training criteria like for example minimum phone error (MPE). This permits to evaluate directly the utility of the margin concept for string recognition. The equivalence relations and the margin-based training criteria lead to a unified view of three major training paradigms, namely Gaussian HMMs, log-linear models, and support vector machines (SVMs). Generalized iterative scaling (GIS) is traditionally used for the optimization of log-linear models with the maximum mutual information (MMI) criterion. This thesis suggests an extension of GIS to log-linear models including hidden variables, and to other training criteria (e.g. MPE). Finally, investigations on convex optimization in speech recognition are presented. Experimental results are provided for a variety of tasks, including the European Parliament plenary sessions task and Mandarin broadcasts.

48 citations


Journal ArticleDOI
TL;DR: A novel human activity recognition method is proposed, which utilizes Independent Component Analysis for activity shape information extraction from image sequences and Hidden Markov Model (HMM) for recognition.
Abstract: In proactive computing, human activity recognition from image sequences is an active research area. In this paper, a novel human activity recognition method is proposed, which utilizes Independent Component Analysis (ICA) for activity shape information extraction from image sequences and Hidden Markov Model (HMM) for recognition. Various human activities are represented by shape feature vectors from the sequence of activity shape images via ICA. Based on these features, each HMM is trained and activity recognition is achieved by the trained HMMs of different activities. Our recognition performance has been compared to the conventional method where Principal Component Analysis (PCA) is typically used to derive activity shape features. Our results show that superior recognition is achieved with the proposed method especially for activities (e.g., skipping) that cannot be easily recognized by the conventional method. Furthermore, by employing Linear Discriminant Analysis (LDA) on IC features, the recognition results further improved significantly in the recognition performance.

47 citations


Proceedings ArticleDOI
23 Aug 2010
TL;DR: Experimental results demonstrate that the presented approach highly accelerates biometric identification and reduces the response time of the system.
Abstract: Performing identification on large-scale biometric databases requires an exhaustive linear search. Since biometric data does not have any natural sorting order, indexing databases, in order to minimize the response time of the system, represents a great challenge. In this work we propose a biometric hash generation technique for the purpose of biometric database indexing, applied to iris biometrics. Experimental results demonstrate that the presented approach highly accelerates biometric identification.

46 citations


Journal ArticleDOI
TL;DR: This paper presents a recognition system for offline signatures using Discrete Cosine Transform (DCT) and Hidden Markov Model (HMM) and shows that successful signatures recognition rates of 99.2% is possible.
Abstract: HMM has been used successfully to model speech and online signature in the past two decades. The success has been attributed to the fact that these biometric traits have time reference. Only few HMM based offline signature recognition systems have be developed because offline signature lack time reference. This paper presents a recognition system for offline signatures using Discrete Cosine Transform (DCT) and Hidden Markov Model (HMM). The signature to be trained or recognized is vertically divided into segments at the centre of gravity using the space reference positions of the pixels. The number of segmented signature blocks is equal to the number of states in the HMM for each user notwithstanding the length of the signatures. Experimental result shows that successful signatures recognition rates of 99.2% is possible. The result is better in comparison with previous related systems based on HMM and statistical classifiers.

41 citations


Journal ArticleDOI
TL;DR: The paper presents a survey of off-line signature verification approaches being followed in different areas and covers some of the examples of the ways these approaches are being followed.
Abstract: Biometrics can be classified into two broad categories— behavioral (signature verification, keystroke dynamics, etc.) and physiological (iris characteristics, fingerprint, etc.). Handwritten signature is amongst the first few biometrics to be used even before the advent of computers. Signature verification is widely studied and discussed using two approaches [5]. On-line approach uses an electronic tablet and a stylus connected to a computer to extract information about a signature and takes dynamic information like; pressure, velocity, etc whereas in offline approach stable dynamic variations are not used for verification purpose. Offline systems are more applicable and easy to use in comparison with on-line systems in many parts of the world however it is considered more difficult than on-line verification due to the lack of dynamic information. The paper presents a survey of off-line signature verification approaches being followed in different areas. This being a nascent area under research, the survey covers some of the examples of the ways

40 citations


Journal ArticleDOI
TL;DR: Issues regarding off-line signature recognitions are discussed, a system designed using cluster based global features which is a multi algorithmic offline signature recognition system is discussed and existing techniques are reviewed.
Abstract: Handwritten signature is one of the most widely used biometric traits for authentication of person as well as document. In this paper we discuss issues regarding off-line signature recognitions. We review existing techniques, their performance and method for feature extraction. We discuss a system designed using cluster based global features which is a multi algorithmic offline signature recognition system.

39 citations


Journal ArticleDOI
TL;DR: Experimental results show that the proposed protected on-line signature recognition system guarantees recognition rates comparable with those of unprotected approaches, and outperforms already proposed protection schemes for signature biometrics.
Abstract: In this paper we propose a biometric cryptosystem able to provide security and renewability to a function based on- line signature representation. A novel reliable signature traits selection procedure, along with a signature binarization algorithm, are introduced. Experimental results, evaluated on the public MCYT signature database, show that the proposed protected on-line signature recognition system guarantees recognition rates comparable with those of unprotected approaches, and outperforms already proposed protection schemes for signature biometrics.

01 Jan 2010
TL;DR: This work establishes that ECG signal is a signature like fingerprint, retinal signature for any individual Identification, and presents a systematic Template matching for Identification of individuals from ECG data.
Abstract: Protection anxiety is to be increased as the technology for forgery grows. Reliable personal Identification and prevention of forged identities is one of the major tasks. Currently, Biometrics is being used extensively for the purpose of security measures. Biometric recognition provides strong security by identifying an individual based on the feature vector(s) derived from their physiological and/or behavioral characteristics. It has been proved that the human Electrocardiogram (ECG) shows adequately unique patterns for biometric recognition. Individual can be identified once ECG signature is formulated. This paper presents a systematic Template matching for Identification of individuals from ECG data. This work establishes that ECG signal is a signature like fingerprint, retinal signature for any individual Identification. Samples of individuals from the MIT/BIH database were taken. The matching decisions are evaluated on the basis of correlation coefficient between features. Preliminary experimental results indicate that the system is accurate (99%), robust, error rate is smaller than 0.9 and achieves a good result for Identification process.

Journal ArticleDOI
TL;DR: This paper incorporates the timing information available in the signature along with the Gabor filter response to generate the feature vector of a dynamic signature.
Abstract: Dynamic signature recognition is one of the commonly used biometric traits. In this paper we propose use of Gabor filters based feature for verification of dynamic signature. We incorporate the timing information available in the signature along with the Gabor filter response to generate the feature vector. Gabor filters have been widely used for image, texture analysis. Here we present analysis for the Gabor filter based feature vector of a dynamic signature.

Patent
29 Jul 2010
TL;DR: A biometric authentication device including a biometric information acquiring unit and a processing unit is described in this paper, where the processing unit extracts, for each block obtained by dividing the biometric input image, a local feature representing the geometric feature of the Biometric input information; classifies the plurality of blocks into a plurality of groups by blocks with a similar local feature; extracts a second group feature representing feature of Biometric Input Information for each group; calculates the degree of difference between each of registered biometrics information and the Biometrics input information based on a first group feature
Abstract: A biometric authentication device including: a biometric information acquiring unit which generates a biometric input image representing user's biometric input information; and a processing unit. The processing unit extracts, for each block obtained by dividing the biometric input image, a local feature representing the geometric feature of the biometric input information; classifies the plurality of blocks into a plurality of groups by blocks with a similar local feature; extracts a second group feature representing feature of biometric input information for each group; calculates the degree of difference between each of registered biometric information and the biometric input information based on a first group feature for each group set for a registered biometric image representing the registered biometric information and the second group feature; selects a prescribed number of registered biometric information based on the degree of the difference; and matches the selected registered biometric information with the biometric input information.

Journal ArticleDOI
TL;DR: In this brief, conic section function neural network (CSFNN) circuitry was designed for offline signature recognition and trained with chip-in-the-loop learning technique in order to compensate typical analog process variations.
Abstract: In this brief, conic section function neural network (CSFNN) circuitry was designed for offline signature recognition. CSFNN is a unified framework for multilayer perceptron (MLP) and radial basis function (RBF) networks to make simultaneous use of advantages of both. The CSFNN circuitry architecture was developed using a mixed mode circuit implementation. The designed circuit system is problem independent. Hence, the general purpose neural network circuit system could be applied to various pattern recognition problems with different network sizes on condition with the maximum network size of 16-16-8. In this brief, CSFNN circuitry system has been applied to two different signature recognition problems. CSFNN circuitry was trained with chip-in-the-loop learning technique in order to compensate typical analog process variations. CSFNN hardware achieved highly comparable computational performances with CSFNN software for nonlinear signature recognition problems.

Journal ArticleDOI
TL;DR: A multimodal biometric approach for identity verification using two competent traits, iris and retina, which diminishes the drawback of single biometric system and improves the performance of an authentication system.

01 Jan 2010
TL;DR: Off-line Signature Verification Based on Fusion of Grid and Global Features Using Neural Networks (SVFGNN) is presented, where the global and grid features are fused to generate set of features for the verification of signature.
Abstract: Signature is widely used and developed area of research for personal verification and authentication. In this paper Off-line Signature Verification Based on Fusion of Grid and Global Features Using Neural Networks (SVFGNN) is presented. The global and grid features are fused to generate set of features for the verification of signature. The test signature is compared with data base signatures based on the set of features and match/non match of signatures is decided with the help of Neural Network. The performance analysis is conducted on random, unskilled and skilled signature forgeries along with genuine signatures. It is observed that FAR and FRR results are improved in the proposed method compared to the existing algorithm. Biometrics is the science of automatic recognition of individual depending on their physiological and behavioral attributes. The expansion of networked society and increased use of some personal portable devices like tablet PCs, PDAs, mobile phones and authorization of access to sensitive data, is demanding the most reliable personal identification and authentication systems. Among the different forms of biometric recognition systems such as fingerprint, iris, face, voice, palm etc., signature will be most widely used. The applications like government and legal financial transaction, bank cheques use signature as one of the personal identification system. The financial transactions and shopping using debit cards and credit cards require a bill to be confirmed by handwritten signature. But this leads to increased risk of financial loss due to attempted forgeries. This problem may be resolved by introducing automatic recognition systems which are being successfully used effectively to analyse large quantities of biometric data. Since olden days handwritten signature has been most widely used and accepted individual attributes for recognition. The design and development of signature recognition system is really big challenge because of the increased dependence of personal identification systems. Signature recognition system is divided into On-line or dynamic and off-line or static recognition. On-line recognition refers to a process where the signer uses a special pen called stylus to create his or her signature, producing the pen locations, speed and pressure, where as off-line recognition deals with signature images acquired by a scanner or a digital camera. In general, off-line signature recognition is a challenging problem, unlike the on-line signature where dynamic aspects of the signing action are captured directly as the handwriting trajectory. Contribution: In this paper, the grid and global features of signature are fused to generate final feature vector of signature. The Neural Network (NN) is used as a classifier for the verification of signatures.

Patent
01 Feb 2010
TL;DR: In this paper, a method of text-based authentication that accounts for positional variability of biometric features between captured biometric data samples is proposed, where the positions of the features are permitted to vary in overlapping border regions within the positional relationship medium.
Abstract: A method of text-based authentication that accounts for positional variability of biometric features between captured biometric data samples includes capturing biometric data for a desired biometric type from an individual, and processing the captured biometric data to generate a biometric image and a biometric feature template. A selected conversion algorithm is executed by superimposing a positional relationship medium on the biometric image. The positional relationship medium includes a plurality of cells textually describable with words derivable from the positional relationship medium. The positions of biometric features are permitted to vary in overlapping border regions within the positional relationship medium. The method also includes identifying the position of at least one biometric feature within the overlapping border regions and generating a plurality of words for the at least one biometric feature.

OtherDOI
27 Dec 2010

Proceedings ArticleDOI
TL;DR: The proposed method is called "signature recognition" because it considers a space-time signature of the behaviour of objects that are used in particular activities (e.g. patients' care in a healthcare environment for elder people with restricted mobility).
Abstract: Automatic estimation of human activities is a topic widely studied. However the process becomes difficult when we want to estimate activities from a video stream, because human activities are dynamic and complex. Furthermore, we have to take into account the amount of information that images provide, since it makes the modelling and estimation activities a hard work. In this paper we propose a method for activity estimation based on object behavior. Objects are located in a delimited observation area and their handling is recorded with a video camera. Activity estimation can be done automatically by analyzing the video sequences. The proposed method is called "signature recognition" because it considers a space-time signature of the behaviour of objects that are used in particular activities (e.g. patients' care in a healthcare environment for elder people with restricted mobility). A pulse is produced when an object appears in or disappears of the observation area. This means there is a change from zero to one or vice versa. These changes are produced by the identification of the objects with a bank of nonlinear correlation filters. Each object is processed independently and produces its own pulses; hence we are able to recognize several objects with different patterns at the same time. The method is applied to estimate three healthcare-related activities of elder people with restricted mobility.

Proceedings ArticleDOI
16 Nov 2010
TL;DR: A new system for dynamic signature verification is presented based on the consideration that each region of an handwritten signature can convey personal characteristics in diverse domains and a multi-expert approach is considered.
Abstract: In this paper a new system for dynamic signature verification is presented. It is based on the consideration that each region of an handwritten signature can convey personal characteristics in diverse domains. Therefore, a multi-expert approach is considered in which each stroke of the signature is evaluated in the most profitable domain of representation. The experimental results demonstrate the effectiveness of the proposed approach.

Journal ArticleDOI
TL;DR: This new method minimizes built in noise of iris images using in-band thresholding in order to provide better mapping and encoding of the relevant information and may provide more flexibility for non-ideal images.
Abstract: —Iris recognition has been demonstrated to be an efficient technology for doing personal identification. In this work, a method to perform iris recognition using biorthogonal wavelets is introduced. Effective use of biorthogonal wavelets using a lifting technique to encode the iris information is demonstrated. This new method minimizes built in noise of iris images using in-band thresholding in order to provide better mapping and encoding of the relevant information. Comparison of Gabor encoding, similar to the method used by Daugman and others, and biorthogonal wavelet encoding is performed. While Daugman's approach is a well-proven algorithm, the effectiveness of our algorithm is shown for the CASIA database, based on the ability to classify inter and intra class distributions, and may provide more flexibility for non-ideal images. The method was tested on the CASIA dataset of iris images with over 4,536 intra-class and 566,244 inter-class comparisons made. After calculating Hamming distances and for the selected threshold value of 0.4, FRR and FAR values were 13.6% and 0.6% using Gabor filter and 0% and 0.03% using the biorthogonal wavelets.

Journal ArticleDOI
TL;DR: It is confirmed that the proposed method exhibits improved recognition accuracy of about 3.6-4.8%, and offers the advantage of lower computational complexity than traditional biometric approaches.

Book ChapterDOI
01 Jan 2010
TL;DR: A hip-based user recognition method which can be suitable for implicit and periodic re-verification of the identity, which uses a wearable accelerometer sensor attached to the hip of the person and the measured hip motion signal is analysed for identity verification purposes.
Abstract: In todays society the demand for reliable verification of a user identity is increasing. Although biometric technologies based on fingerprint or iris can provide accurate and reliable recognition performance, they are inconvenient for periodic or frequent re-verification. In this paper we propose a hip-based user recognition method which can be suitable for implicit and periodic re-verification of the identity. In our approach we use a wearable accelerometer sensor attached to the hip of the person, and then the measured hip motion signal is analysed for identity verification purposes. The main analyses steps consists of detecting gait cycles in the signal and matching two sets of detected gait cycles. Evaluating the approach on a hip data set consisting of 400 gait sequences (samples) from 100 subjects, we obtained equal error rate (EER) of 7.5% and identification rate at rank 1 was 81.4%. These numbers are improvements by 37.5% and 11.2% respectively of the previous study using the same data set.

Patent
08 Apr 2010
TL;DR: In this paper, a method for validating an individual's identity by using a combination of a RFID card, signature captured by a touch screen computer, and a facial photo captured by the built-in camera of a touch-screen computer is provided.
Abstract: A method is provided for validating an individual's identity by using a combination of a RFID card, signature captured by a touch screen computer, and a facial photo captured by a built-in camera of a touch screen computer. Using third party biometric signature recognition algorithms and facial recognition algorithms are preferred to differentiate the differences between a sample signature and facial photo referenced by a RFID card value.

Patent
05 Nov 2010
TL;DR: In this paper, a system and method for generating an encryption key using physical characteristics of a biometric sample is described, where the biometric feature(s) from a sample are analyzed to generate a feature vector.
Abstract: A system and method for generating an encryption key using physical characteristics of a biometric sample is described. In one embodiment, the biometric feature(s) from a sample are analyzed to generate a feature vector. After discretizing the feature(s), the resultant feature vector is translated into a bit vector. The bit vector is the secure biometric key that results from the biometric(s). The secure biometric key is used to generate at least one cryptographic key. A similar process is used to access the cryptographic key secured by the secure biometric key. If the access biometric key matches the secure biometric key, the cryptographic key is revealed and access is allowed. In another embodiment, if the access biometric key does not match the secure biometric key a camouflaging process is used to provide an unauthorized user a bogus secure biometric key indistinguishable from the correct secure biometric key.

Proceedings ArticleDOI
28 May 2010
TL;DR: In this work, the pen-position parameters of the online signature are decomposed into multiscale signals by using the wavelet transform technique and a TESPAR DZ based method is employed to code the approximation and details coefficients.
Abstract: In this work, a new on-line signature verification system is proposed. Firstly, the pen-position parameters of the online signature are decomposed into multiscale signals by using the wavelet transform technique. A TESPAR DZ based method is employed to code the approximation and details coefficients. Thus, for each analyzed time function, a fixed dimension feature vector is obtained. Experimental results were reported using the SVC2004 database. The models were trained and tested with the Support Vector Machine classifier. A feature level fusion strategy was adapted.

Proceedings ArticleDOI
13 Apr 2010
TL;DR: Algorithms to reliably generate biometric identifiers from a user's biometric image which in turn is used for identity verification possibly in conjunction with cryptographic keys, and assures privacy of the biometric using the one-way hashing property.
Abstract: We present algorithms to reliably generate biometric identifiers from a user's biometric image which in turn is used for identity verification possibly in conjunction with cryptographic keys. The biometric identifier generation algorithms employ image hashing functions using singular value decomposition and support vector classification techniques. Our algorithms capture generic biometric features that ensure unique and repeatable biometric identifiers. We provide an empirical evaluation of our techniques using 2569 images of 488 different individuals for three types of biometric images; namely fingerprint, iris and face. Based on the biometric type and the classification models, as a result of the empirical evaluation we can generate biometric identifiers ranging from 64 bits up to 214 bits. We provide an example use of the biometric identifiers in privacy preserving multi-factor identity verification based on zero knowledge proofs. Therefore several identity verification factors, including various traditional identity attributes, can be used in conjunction with one or more biometrics of the individual to provide strong identity verification. We also ensure security and privacy of the biometric data. More specifically, we analyze several attack scenarios. We assure privacy of the biometric using the one-way hashing property, in that no information about the original biometric image is revealed from the biometric identifier.

Patent
Aaron K. Baughman1
15 Jul 2010
TL;DR: In this paper, a system, method and program product for generating a private key was described, which includes a signal acquisition system for obtaining biometric input from a user and encoding the biometric inputs into an acquired biometric; a recognition system for determining an identity based on the acquired biometrics and outputting an absolute biometric associated with the identity.
Abstract: A system, method and program product for generating a private key. A system is disclosed that includes a signal acquisition system for obtaining biometric input from a user and encoding the biometric input into an acquired biometric; a recognition system for determining an identity based on the acquired biometric and outputting an absolute biometric associated with the identity; an input device for accepting a knowledge input from the user; and a key generator that generates a private key based on the knowledge input and the absolute biometric.

Journal ArticleDOI
TL;DR: The proposed system aimed to provide simple, faster robust system using less number of features when compared to state of art works and showed that the SVM classifier yielded the most promising 8% False Rejection Rate (FRR) and 10% False acceptance Rate (FAR).
Abstract: Problem statement: The research addressed the computational load reduction in off-line signature verification based on minimal features using bayes classifier, fast Fourier transform, linear discriminant analysis, principal component analysis and support vector machine approaches. Approach: The variation of signature in genuine cases is studied extensively, to predict the set of quad tree components in a genuine sample for one person with minimum variance criteria. Using training samples, with a high degree of certainty the Minimum Variance Quad tree Components (MVQC) of a signature for a person are listed to apply on imposter sample. First, Hu moment is applied on the selected subsections. The summation values of the subsections are provided as feature to classifiers. Results: Results showed that the SVM classifier yielded the most promising 8% False Rejection Rate (FRR) and 10% False Acceptance Rate (FAR). The signature is a biometric, where variations in a genuine case, is a natural expectation. In the genuine signature, certain parts of signature vary from one instance to another. Conclusion: The proposed system aimed to provide simple, faster robust system using less number of features when compared to state of art works.