Large image generation and vision models, combined with differentiable rendering technologies, have become powerful tools for generating paths that can be drawn or painted by a robot. However, these tools often overlook the intrinsic physicality of the human drawing/writing act, which is usually executed with skillful hand/arm gestures. Taking this into account is important for the visual aesthetics of the results and for the development of closer and more intuitive artist-robot collaboration scenarios. We present a method that bridges this gap by enabling gradient-based optimization of natural human-like motions guided by cost functions defined in image space. To this end, we use the sigma-lognormal model of human hand/arm movements, with an adaptation that enables its use in conjunction with a differentiable vector graphics (DiffVG) renderer. We demonstrate how this pipeline can be used to generate feasible trajectories for a robot by combining image-driven objectives with a minimum-time smoothing criterion. We demonstrate applications with generation and robotic reproduction of synthetic graffiti as well as image abstraction.

Handwriting movements can be leveraged as a unique form of behavioral biometrics, to verify whether a real user is operating a device or application. This task can be framed as a “reverse Turing test” in which a computer has to detect if an input instance has been generated by a human or artificially. To tackle this task, we study ten public datasets of handwritten symbols (isolated characters, digits, gestures, pointing traces, and signatures) that are artificially reproduced using seven different synthesizers, including, among others, the Kinematic Theory (ΣΛ model), generative adversarial networks, Transformers, and Diffusion models. We train a shallow recurrent neural network that achieves excellent performance (98.3% Area Under the ROC Curve (AUC) score and 1.4% equal error rate on average across all synthesizers and datasets) using nonfeaturized trajectory data as input. In few-shot settings, we show that our classifier achieves such an excellent performance when trained on just 10% of the data, as evaluated on the remaining 90% of the data as a test set. We further challenge our classifier in out-of-domain settings, and observe very competitive results as well. Our work has implications for computerized systems that need to verify human presence, and adds an additional layer of security to keep attackers at bay.

Signature synthesis is a computation technique that generates artificial specimens which can support decision making in automatic signature verification. A lot of work has been dedicated to this subject, which centres on synthesizing dynamic and static two-dimensional handwriting on canvas. This paper proposes a framework to generate synthetic 3D on-air signatures exploiting the lognormality principle, which mimics the complex neuromotor control processes at play as the fingertip moves. Addressing the usual cases involving the development of artificial individuals and duplicated samples, this paper contributes to the synthesis of: (1) the trajectory and velocity of entirely 3D new signatures; (2) kinematic information when only the 3D trajectory of the signature is known, and (3) duplicate samples of 3D real signatures. Validation was conducted by generating synthetic 3D signature databases mimicking real ones and showing that automatic signature verifications of genuine and skilled forgeries report performances similar to those of real and synthetic databases. We also observed that training 3D automatic signature verifiers with duplicates can reduce errors. We further demonstrated that our proposal is also valid for synthesizing 3D air writing and gestures. Finally, a perception test confirmed the human likeness of the generated specimens. The databases generated are publicly available, only for research purposes, at .

The Kinematic Theory of rapid movements, and its associated Sigma-Lognormal, model 2D spatiotemporal trajectories. It is constructed mainly as a temporal overlap of curves between virtual target points. Specifically, it uses an arc and a lognormal as primitives for the representation of the trajectory and velocity, respectively. This paper proposes developing this model, in what we call the Kinematic Theory Transform, which establishes a mathematical framework that allows further primitives to be used. Mainly, we evaluate Euler curves to link virtual target points and Gaussian, Beta, Gamma, Double-bounded lognormal, and Generalized Extreme Value functions to model the bell-shaped velocity profile. Using these primitives, we report reconstruction results with spatiotemporal trajectories executed by human beings, animals, and anthropomorphic robots.

This paper proposes a method to generate the synthetic kinematic of signatures in 3D. The analysis of 3D signatures is becoming a hot topic due to the irruption of commercial off-the-shelf devices for easy acquisition of 3D movements. However, the novelty of this technology reveals the scarce publicly available signatures in 3D, which hinder their de-velopment. A solution is the synthesis of Signatures in 3D. As a first step, this paper synthesizes the kinematics of 3D signatures based on the Kinematic Theory of Rapid Movements and its associated Sigma-Lognormal model in 3D. To evaluate the method, we regenerate signature databases with synthetic speed profiles in all genuine and forgeries found in two 3D signature databases. Then, we analyze the similarities in the performance of a signature verifier when real and synthetic signatures are used in random and skilled forgeries experiments.

Motor skills are fundamental for the development of children. Neurodevelopmental tests currently used by professionals for measuring motor control maturity exhibit several limitations. To address some of these, we have designed the Lognometer, a tablet-based device that can run computerized neuromotor tests. To normalize this tool against a representative population, we collected handwritten triangles from 780 children. We used the Sigma-Lognormal model and a prototype-based parameter estimation algorithm to analyze these movements. To ensure clinical acceptance, we developed an explainable solution relying on statistical regression. We evaluated how well the proposed lognormal decomposition captures the motor control maturation between 6 to 13 years of age by plotting the biological age versus the age estimated using movement kinematics. To provide an equivalent to growth curves, we further overlaid percentile lines that can be used by clinicians to evaluate the neuromotor development of children.

It is important to accurately evaluate the motor control maturity to help physicians diagnose delayed or abnormal motor development in children. Traditionally, it has been challenging to design assessment methods that are practical and accurate at the same time. This study aims to develop an effective algorithm to predict motor control maturity based on the Kinematic Theory of rapid human movements. We used handwritten pen strokes made on an electronic tablet by 513 children (5.5 to 13 years of age). We considered two types of movements: a single stroke and a triangle drawing test. For the analysis, Sigma-Lognormal parameters were extracted from recordings and used in predictive models. We compared multiple models, including linear regression, deep learning, K-nearest neighbors regression, random forest, gradient boosting regression, and support vector regression. These models performed well considering the within-children variability in handwritten strokes and the between-children variability in motor control maturity. The best score was obtained using the neural network model: coefficient of determination (R2): 0.548; mean absolute error (MAE): 0.937. We found simple stroke parameters alone to be sub-optimal; the results were better when using parameters from the triangle test. In conclusion, our study demonstrates that the Sigma-Lognormal model offers new possibilities for estimating the motor control maturity. Our method is fast and comfortable for the children, as it only requires performing handwriting strokes on an electronic tablet. This simple and user-friendly test is expected to be more convenient for doctors in a clinical context.

Muscle force profiles depend on the sequential activation of motor units. Actual models are all phenomenological: they are based on the Hill-type model of muscle actuation. This work presents an original and innovative approach to produce muscle activation profiles from the proportionality hypothesis of the Kinematic Theory of Human Movement. The Kinematic Theory makes some of the most accurate predictions of human velocity profiles for a large variety of movements as the impulse response of a complex system. We demonstrate that the cumulative time delay derived from polygonal distributions of muscle fibers will always converge and that it is possible to define the limit with only two parameters. These models are combined with an exponential deactivation model that, together, accurately reproduce experimental electromyographic profiles. In the mid-term, it is expected that this type of model will open brand new applications in handwriting recognition.

Online fraud often involves identity theft. Since most security measures are weak or can be spoofed, we investigate a more nuanced and less explored avenue: behavioral biometrics via handwriting movements. This kind of data can be used to verify whether a user is operating a device or a computer application, so it is important to distinguish between human and machine-generated movements reliably. For this purpose, we study handwritten symbols (isolated characters, digits, gestures, and signatures) produced by humans and machines, and compare and contrast several deep learning models. We find that if symbols are presented as static images, they can fool state-of-the-art classifiers (near 75% accuracy in the best case) but can be distinguished with remarkable accuracy if they are presented as temporal sequences (95% accuracy in the average case). We conclude that an accurate detection of fake movements has more to do with how users write, rather than what they write. Our work has implications for computerized systems that need to authenticate or verify legitimate human users, and provides an additional layer of security to keep attackers at bay.

The Kinematic Theory of rapid movements and its associated Sigma-Lognormal model have been extensively used in a large variety of applications. While the physical and biological meaning of the model have been widely tested and validated for rapid movements, some shortcomings have been detected when it is used with continuous long and complex movements. To alleviate such drawbacks, and inspired by the motor equivalence theory and a conceivable visual feedback, this paper proposes a novel framework to extract the Sigma-Lognormal parameters, namely iDeLog. Specifically, iDeLog consists of two steps. The first one, influenced by the motor equivalence model, separately derives an initial action plan defined by a set of virtual points and angles from the trajectory and a sequence of lognormals from the velocity. In the second step, based on a hypothetical visual feedback compatible with an open-loop motor control, the virtual target points of the action plan are iteratively moved to improve the matching between the observed and reconstructed trajectory and velocity. During experiments conducted with handwritten signatures, iDeLog obtained promising results as compared to the previous development of the Sigma-Lognormal.

We introduce KeyTime, a new technique and accompanying software for predicting the production times of users' stroke gestures articulated on touchscreens. KeyTime employs the principles and concepts of the Kinematic Theory, such as lognormal modeling of stroke gestures' velocity profiles, to estimate gesture production times significantly more accurately than existing approaches. Our experimental results obtained on several public datasets show that KeyTime predicts user-independent production times that correlate r=.99 with groundtruth from just one example of a gesture articulation, while delivering an average error in the predicted time magnitude that is 3 to 6 times smaller than that delivered by CLC, the best prediction technique up to date. Moreover, KeyTime reports a wide range of useful statistics, such as the trimmed mean, median, standard deviation, and confidence intervals, providing practitioners with unprecedented levels of accuracy and sophistication to characterize their users' a priori time performance with stroke gesture input.

Handwriting is a fundamental skill that each pupil should master for successfully completing his instructions. However, for a certain number of children, after a successful phase of handwriting learning an unexplained rough deterioration of their efficiency on paper is observed during the phase of customization of their handwriting. In this context, our goal is to make a comprehensive study of the evolution and the difficulties of children handwriting learning according to handwriting teaching approaches involved in school. To achieve such a challenge, it is necessary to collect in a secured way and to analyze a large amount of various contextualized online and offline handwritten data produced in real scholar situations by numerous pupils from kindergarten up to middle school. This is the purpose of the ongoing @MaGma project that was defined with the Academic direction of the Guadeloupian Region. In this paper, we specify the problems handled in @Magma and depict the general principles which will have to govern the collaborative infrastructure of acquisition and treatment of children's writing, based on 2 frameworks: Copilotr@ce and Dekattras. Next, we report the preliminary results of the comparative sigma-lognormal and dynamic analysis of a set of children scribbles acquired thanks to this infrastructure. We conclude by developing how these first results obtained in a real scholar acquisition context confirm the experimental results previously obtained in more clinical contexts, pointing out the fact that the Personalized Digital Bodyguard concept and vision is realizable.

The dynamic signature is a biometric trait widely used and accepted for verifying a person's identity. Current automatic signature-based biometric systems typically require five, ten, or even more specimens of a person's signature to learn intrapersonal variability sufficient to provide an accurate verification of the individual's identity. To mitigate this drawback, this paper proposes a procedure for training with only a single reference signature. Our strategy consists of duplicating the given signature a number of times and training an automatic signature verifier with each of the resulting signatures. The duplication scheme is based on a sigma lognormal decomposition of the reference signature. Two methods are presented to create human-like duplicated signatures: the first varies the strokes' lognormal parameters (stroke-wise) whereas the second modifies their virtual target points (target-wise). A challenging benchmark, assessed with multiple state-of-the-art automatic signature verifiers and multiple databases, proves the robustness of the system. Experimental results suggest that our system, with a single reference signature, is capable of achieving a similar performance to standard verifiers trained with up to five signature specimens.

We show that the Kinematic Theory produces synthesized stroke gestures that ‘look and feel’ the same and hold the same statistical characteristics as human-generated gestures. Previous research in this vein has conducted such comparison from the classification accuracy performance, which is a legitimate though indirect measure. In this article, we synthesized two well-known public data sets comprising unistroke and multistroke gestures. We then compared geometric, kinematic and articulation aspects of human and synthetic gestures, and found no practical differences between both populations. We also conducted an online survey involving 236 participants and found that it is very difficult to tell human and synthetic gestures apart. We can finally be confident that synthesized gestures are actually reflective of how users produce stroke gestures. In sum, this work enables a deeper understanding of synthetic gestures’ production, which can inform the design of better gesture sets and development of more accurate recognizers.

When using tablet computers, smartphones, or digital pens, human users perform movements with a stylus or their fingers that can be analyzed by the kinematic theory of rapid human movements. In this paper, we present a user-centered system for signature verification that performs such a kinematic analysis to verify the identity of the user. It is one of the first systems that is based on a direct comparison of the elementary neuromuscular strokes which are detected in the handwriting. Taking into account the number of strokes, their similarity, and their timing, the string edit distance is employed to derive a dissimilarity measure for signature verification. On several benchmark datasets, we demonstrate that this neuromuscular analysis is complementary to a well-established verification using dynamic time warping. By combining both approaches, our verifier is able to outperform current state-of-the-art results in on-line signature verification.

The papers in this special section focus on handwriting and drawing processes for user-centered systems. The papers provide a wide and updated overview of the frontier of research in the field of humancentered systems based on drawing and handwriting processing. Through the papers, some of the most relevant directions of further research are highlighted with specific attention to components related to human–machine interaction. The Guest Editors hope that this issue brings forth the importance of automated systems related to automatic processing of drawing and handwriting.

Handheld touch-capable devices have become one of the most popular and fastest growing consumer products. It seems logical therefore to think of such devices as Personal Digital Bodyguards (PDBs) in charge for example of biometrical, biomedical, and neurocognitive monitoring by just inspecting the user's handwriting activity. However, it is unclear whether the hardware of today's devices is capable to handle this task. To this end, we conducted a comparative study regarding the capabilities of past and current tablets to allow for the design of PDBs based on the exploitation of the Kinematic Theory. Our study shows that, while some improvements are still necessary at the sampling frequency level, the conclusions drawn by the Kinematic Theory can be directly transferred to PDBs.

In this paper, the existence of periodic and partly periodic oscillation for a recurrent neural network with time-varying input and time delays between neural interconnections is investigated. Some theorems to determine the conditions for periodic oscillations are demonstrated. Simple and practical criteria for selecting the parameters in this network are derived. Typical simulation examples are also presented to illustrate the whole methodology.

Training a high-quality gesture recognizer requires providing a large number of examples to enable good performance on unseen, future data. However, recruiting participants, data collection, and labeling, etc., necessary for achieving this goal are usually time consuming and expensive. Thus, it is important to investigate how to empower developers to quickly collect gesture samples for improving UI usage and user experience. In response to this need, we introduce Gestures à Go Go ( g 3), a web service plus an accompanying web application for bootstrapping stroke gesture samples based on the kinematic theory of rapid human movements. The user only has to provide a gesture example once, and g 3 will create a model of that gesture. Then, by introducing local and global perturbations to the model parameters, g 3 generates from tens to thousands of synthetic human-like samples. Through a comprehensive evaluation, we show that synthesized gestures perform equally similar to gestures generated by human users. Ultimately, this work informs our understanding of designing better user interfaces that are driven by gestures.

What can be done with only one enrolled real hand-written signature in Automatic Signature Verification (ASV)? Using 5 or 10 signatures for training is the most common case to evaluate ASV. In the scarcely addressed case of only one available signature for training, we propose to use modified duplicates. Our novel technique relies on a fully neuromuscular representation of the signatures based on the Kinematic Theory of rapid human movements and its Sigma-Lognormal model. This way, a real on-line signature is converted into the Sigma-Lognormal model domain. The model parameters are then varied to generate new duplicated signatures.

In the field of automatic signature verification, a major challenge for statistical analysis and pattern recognition is the small number of reference signatures per user. Score normalization, in particular, is challenged by the lack of information about intra-user variability. In this paper, we analyze several approaches to score normalization for dynamic time warping and propose a new two-stage normalization which detects simple forgeries in a first stage and copes with more skilled forgeries in a second stage. An experimental evaluation is conducted on two data sets with different characteristics, namely the MCYT online signature corpus, which contains over three hundred users, and the SUSIG visual sub-corpus, which contains highly skilled forgeries. The results demonstrate that score normalization is a key component for signature verification and that the proposed two-stage normalization achieves some of the best results on these difficult data sets both for random and for skilled forgeries.

A fully automatic framework based on the kinematic theory of rapid human movements was recently introduced for analyzing and modeling complex human movements patterns such as those involved in handwriting. In this paper, we present a new approach to better extract and estimate the lognormal primitives and parameters. Through a comprehensive evaluation using 32,000 words from a public database, we show that our approach greatly improves the state-of-the-art extractor.

Word level handwritten CAPTCHA generation involves picking a handwritten word from a pre-existing database and cumulatively applying distortions and noise models. In principle, the addition of distortion and noise makes the CAPTCHA robust to automated attacks. However, the primary drawback of the word level CAPTCHA generation is that it limits us to words that already exist in our data set. If the primary building block of this approach was a character, we could move away from a lexicon based CAPTCHA generation and generate CAPTCHAs which are resistant to a dictionary based attack. In this paper, we propose a Sigma-Lognormal based approach to generate character level CAPTCHAs. Next, we increase the robustness of the model by applying ideas from accents in handwriting to our problem. Finally, we demonstrate the efficacy of our approach by simulating an attack by an automated word recognizer.

A fully automatic framework has been introduced recently for neuromuscular representation of complex handwriting patterns, such as gestures, signatures, and words, based on the Kinematic Theory of rapid human movements and its Sigma-Lognormal model. In this paper, we investigate the application of this framework to unconstrained whiteboard notes, taking into account a novel acquisition modality, multiple writers, natural language, and complete text lines. Although these conditions deviate strongly from the previously considered scenario of brief pen movements on tablet computers, we demonstrate that the Sigma-Lognormal model is still able to represent the handwriting accurately. In order to deal with longer handwriting patterns, we propose a robust component-wise representation of text lines that achieves a high model quality. Furthermore, we propose a stroke-wise distortion method to generate synthetic text lines from the Sigma-Lognormal representation of real specimens. For handwriting recognition on the IAM online database, it is demonstrated that the extension of the training set with the proposed synthesis method significantly increases current benchmark results achieved with recurrent neural networks.

A method for automatic generation of synthetic handwritten words is presented which is based in the Kinematic Theory and its Sigma-lognormal model. To generate a new synthetic sample, first a real word is modelled using the Sigma-lognormal model. Then the Sigma-lognormal parameters are randomly perturbed within a range, introducing human-like variations in the sample. Finally, the velocity function is recalculated taking into account the new parameters. The synthetic words are then used as training data for a Hidden Markov Model based on-line handwritten recognizer. The experimental results confirm the great potential of the kinematic theory of rapid human movements applied to writer adaptation.

Popular CAPTCHA systems consist of garbled printed text character images with significant distortions and noise. It is believed that humans have little difficulty in deciphering the text, whereas automated systems are foiled by the added noise and distortion. However, in recent years, several text based CAPTCHAs have been reported as broken, that is, automated systems can identify the text in the displayed image with a reasonable amount of success. An extension to the text based CAPTCHA concept is to utilize unconstrained handwritten text, which is still considered to be a challenging problem for automated systems. In this work, we present a automated handwritten CAPTCHA generation system by adding distortions to the Sigma-Lognormal representation of a handwritten word sample. In addition, several noise models are also considered. We perform experiments on the UNIPEN dataset and demonstrate the efficacy of the approach.

In this paper, the existence of oscillatory solutions for a four coupled FHN network model with delays is investigated. Some theorems to determine the oscillatory solutions for the system are obtained. The practical criteria for selecting the parameters in this network are provided. Computer simulations are also given to illustrate the effectiveness of the results.

Recently, research in handwritten signature verification has been considered with renewed interest. In fact, in the age of e-society, handwritten signature still represents an extraordinary means for personal verification and the possibility of using automatic signature verification in a range of applications is becoming a reality. This paper focuses on some of the most remarkable aspects the field and highlights some recent research directions. A list of selected publications is also provided for interested researchers.

This keynote lecture is divided in two parts. On the one hand, it explores how the modelling of pen strokes can be exploited to design biomedical tools for the analysis of neuromuscular systems to be used for the development of a diagnostic protocol for the assessment of brain stroke risk factors. On the other hand, it explains how the methodology used to model a neuromuscular system producing handwriting strokes can be generalized through various strides, to model the Solar System, the Milky Way and the whole Universe.

Fast reaching movements are an important component of our daily interaction with the world and are consequently under investigation in many fields of science and engineering. Today, useful models are available for such studies, with tools for solving the inverse dynamics problem involved by these analyses. These tools generally provide a set of model parameters that allows an accurate and locally optimal reconstruction of the original movements. Although the solutions that they generate may provide a data curve fitting that is sufficient for some pattern recognition applications, the best possible solution is often necessary in others, particularly those involving neuroscience and biomedical signal processing. To generate these solutions, we present a globally optimal parameter extractor for the delta-lognormal modeling of reaching movements based on the branch-and-bound strategy. This algorithm is used to test the impact of white noise on the delta-lognormal modeling of reaching movements and to benchmark the state-of-the-art locally optimal algorithm. Our study shows that, even with globally optimal solutions, parameter averaging is important for obtaining reliable figures. It concludes that physiologically derived rules are necessary, in addition to global optimality, to achieve meaningful ∆Λ extractions which can be used to investigate the control patterns of these movement primitives.

An analysis of the quality of on-line handwritten signatures is carried out based on the Sigma-Lognormal model. In the study, two main issues are addressed from a kinematic perspective of humanly-produced movements. On the one hand, what makes some signatures perform better than others in automatic signature verification systems, and on the other hand if that information may be used as a quality measure in order to predict the expected performance of a given sample. Experiments were carried out on the MCYT database and show the high potential of certain kinematic features for signature quality assessment.

The kinematical information present in synthetically generated signatures is analyzed using the Sigma-Lognormal model and compared to the kinematical properties of real samples. Experiments are carried out on totally independent development and test sets and show a high degree of similarity between humanly produced and artificial signatures. One particular flaw is found in the velocity profile of synthetic signatures. Two possible solutions are proposed to improve the synthetic generation method using the Kinematic Theory of rapid human movements.

Summary form only. Many models have been proposed over the years to study human movements in general and handwriting in particular: models relying on neural networks, dynamics models, psychophysical models, kinematic models and models exploiting minimization principles. Among the models that can be used to provide analytical representations of a pen stroke, the Kinematic Theory of rapid human movements and its delta-lognormal model has often served as a guide in the design of pattern recognition systems relying on the exploitation of the fine neuromotricity, like on-line handwriting recognition, signature verification as well as in the design of intelligent systems involving in a way or another, the global processing of human movements. Among other things, this invited lecture aims at elaborating a theoretical background for many handwriting applications as well as providing some basic knowledge that could be integrated or taking care of in the development of automatic pattern recognition systems. More specifically, we will overview the basic neuromotor properties of single strokes and explain how they can be superimposed vectorially to generate complex pen tip trajectories. Doing so, we will report on various projects conducted by our team and our collaborators. First, we will present a brief comparative survey of the different models in the field and focus on the family of models involving lognormal functions. Then, from a practical perspective, we will describe two new parameter extraction algorithms suitable for the reverse engineering of individual strokes as well as of complex handwriting signals. We will show how the resulting representation could be employed to characterize signers and writers and how the corresponding feature sets could be exploited to study the effects of various factors, like aging and health problems, on handwriting variability. We will also describe some methodologies to generate automatically huge on-line handwriting databases for either writer dependent or writer independent applications as well as for the production of synthetic signature databases. From a theoretical perspective, we will explain how, using an original psychophysical set up, we have been able to validate the basic hypothesis of the Kinematic Theory and to test its most distinctive predictions. We will complete this survey by explaining how the Kinematic Theory could be utilized to improve electromyographic and electroencephalographic signal processing, opening a window on novel potential applications for on-line handwriting processing, particularly in biomedical engineering and in some fields of the neurosciences.

A three steps methodology is proposed to derive consistent sets of local features which may be easily compared between the different samples of a stereotypical human handwriting movement, allowing the statistical analysis its local variability. This technique is illustrated using the Sigma-Lognormal modeling of on-line triangular trajectory patterns obtained from a standardized neuromuscular task. The overall approach can be adapted and generalized to the analysis of the end-effector kinematics of many planar upper limb movements.

In this paper, the existence of oscillations for a class of recurrent neural networks with time delays between neural interconnections is investigated. By using the fixed point theory and Liapunov functional, we prove that a recurrent neural network might have a unique equilibrium point which is unstable. This particular type of instability, combined with the boundedness of the solutions of the system, will force the network to generate a permanent oscillation. Some necessary and sufficient conditions for these oscillations are obtained. Simple and practical criteria for fixing the range of parameters in this network are also derived. Typical simulation examples are presented.

In this paper, we present a new analytical method for estimating the parameters of Delta-Lognormal functions and characterizing handwriting strokes. According to the Kinematic Theory of rapid human movements, these parameters contain information on both the motor commands and the timing properties of a neuromuscular system. The new algorithm, called XZERO, exploits relationships between the zero crossings of the first and second time derivatives of a lognormal function and its four basic parameters. The methodology is described and then evaluated under various testing conditions. The new tool allows a greater variety of stroke patterns to be processed automatically. Furthermore, for the first time, the extraction accuracy is quantified empirically, taking advantage of the exponential relationships that link the dispersion of the extraction errors with its signal-to-noise ratio. A new extraction system which combines this algorithm with two other previously published methods is also described and evaluated. This system provides researchers involved in various domains of pattern analysis and artificial intelligence with new tools for the basic study of single strokes as primitives for understanding rapid human movements.

The Sigma-Lognormal model of the kinematic theory of rapid human movements, has been implemented in an interactive software tool, allowing the generation of databases of unlimited size from a few online handwriting specimens of letters and words. Online trajectories of a target word produced by a few writers are fitted by the Sigma-Lognormal parameters; using the interactive system. Thereafter, the fiducial pattern of the word is constructed and the writer variability is circumscribed respectively from the mean values and the standard deviations of the extracted parameters. Typical simulation results obtained by randomly fixing the parameters inside these realistic intervals are presented to highlight the ability of the generator to produce a large variety of multi-writer and writer-dependent handwriting patterns as observed in real data. Overall, this software tool provides new insights on the development of huge databases for the training and testing of online handwriting classifiers and recognizers.

A theory, called the Kinematic Theory of Rapid Human Movement, was proposed a few years ago to analyze rapid human movements, called the Kinematic Theory of Rapid Human Movements, based on a delta-lognormal equation that globally describes the basic properties of the velocity profiles of an end-effector using seven parameters. This realistic model has been very useful for proposing original solutions to various pattern recognition problems (signature segmentation and verification, handwriting analysis and synthesis, etc.). Most of these applications rely on the use of an efficient algorithm to extract the delta-lognormal parameters from real data with the best possible fit. In this paper, we compare two such algorithms: a deterministic one, based on nonlinear regression, and a Breeder Genetic algorithm. The performance of these two algorithms and of their combinations are compared using the same artificial database, composed of analytical delta-lognormal profiles and their noisy versions (20 dB SNR). In the free-noise case, the analysis of the experimental results shows that the deterministic approach leads to better results than the evolutionary one, while under the extremely noisy conditions selected, the evolutionary approach seems to be less sensitive to noise, but is nevertheless less successful than the deterministic search.

A few years ago, a Kinematic Theory was proposed to analyze rapid human movements. The theory is based on a delta-lognormal equation which can be used to globally describe the basic properties of velocity profiles using seven parameters. This realistic model has been of great use to solve pattern recognition problems (signature verification, handwriting analysis and segmentation, etc.). To go further in that direction, a better understanding of the model is a prerequisite. This can be either in the context of psychophysical studies involving human subjects or in the context of computer simulations. In this paper, we use the same model form to develop a simulator that generates human-like velocity profiles. A basic subsystem model is both proposed and constructed with a Simulink Matlab tool; then many of these are connected to create an artificial neuromuscular network. Combining two networks in parallel, one agonist and the other antagonist, a synergy simulator is constructed. The similarity of the velocity patterns produced by the simulator is analyzed using a delta-lognormal parameter extractor. It is shown that the parameters extracted from artificially generated profiles vary in the same intervals as those of experimental profiles produced by human subjects. In future works the simulator tool will be used to study the control of rapid human movements.

By using the method of Liapunov functional, a model for bidirectional associative memory networks with time delays is studied. The asymptotic stability is global in the state space of the neuronal activations and is also independent of the delays. Our results can be applied to a variety of situations that arise both in the field of biological and artificial neural networks.

Hidden Markov models (HMM) are stochastic models capable of statistical learning and classification. They have been applied in speech recognition and handwriting recognition because of their great adaptability and versatility in handling sequential signals. On the other hand, as these models have a complex structure and also because the involved data sets usually contain uncertainty, it is difficult to analyze the multiple observation training problem without certain assumptions. For many years researchers have used the training equations of Levinson (1983) in speech and handwriting applications, simply assuming that all observations are independent of each other. This paper presents a formal treatment of HMM multiple observation training without imposing the above assumption. In this treatment, the multiple observation probability is expressed as a combination of individual observation probabilities without losing generality. This combinatorial method gives one more freedom in making different dependence-independence assumptions. By generalizing Baum's auxiliary function into this framework and building up an associated objective function using the Lagrange multiplier method, it is proven that the derived training equations guarantee the maximization of the objective function. Furthermore, we show that Levinson's training equations can be easily derived as a special case in this treatment.

Handwriting has continued to persist as a means of communication and recording information in day-to-day life even with the introduction of new technologies. Given its ubiquity in human transactions, machine recognition of handwriting has practical significance, as in reading handwritten notes in a PDA, in postal addresses on envelopes, in amounts in bank checks, in handwritten fields in forms, etc. This overview describes the nature of handwritten language, how it is transduced into electronic data, and the basic concepts behind written language recognition algorithms. Both the online case (which pertains to the availability of trajectory data during writing) and the off-line case (which pertains to scanned images) are considered. Algorithms for preprocessing, character and word recognition, and performance with practical systems are indicated. Other fields of application, like signature verification, writer authentification, handwriting learning tools are also considered.

This paper presents a brief overview of several technical issues concerning security and commercial aspects of the electronic handling and transfer of documents. After introducing a few basic notions and general definitions, we present a new digital envelope transfer protocol-based system, named TRANZIX, used in the context of generic "transport and transfer of value". We present the general concept of the innovative application that allows commercial exchange of electronic documents through digital transactions using customized electronic money. The system provides very strong protection based on public-key cryptography complemented with biometric protection based on handwritten signatures. Finally, a few targeted applications are overviewed, including the customized exchanges of documents through "mixed transfers of value".

This paper presents a segmentation method that partly mimics the cognitive-behavioral process used by human subjects to recover motor-temporal information from the image of a handwritten word. The approach does not exploit any thinning or skeletonization procedure, but rather a different type of information is manipulated concerning the curvature function of the word contour. In this way, it is possible to detect the parts of the image where the original odometric information is lost or ambiguous (such as, for example, at an intersection of the handwritten lines) and interpret them to finally recover a part of the original temporal information. The algorithm scans the word, following the natural course of the line, and attempts to reproduce the same movement as executed by the writer during the generation of the word. It segments the cursive trace where the contour shows the slow-down of the original movement (corresponding to the maximum curvature points of the curve). At the end of the scanning process, a temporal sequence of motor strokes is obtained which plausibly composed the original intended movement.

Presents a hidden Markov model (HMM) based approach to online handwritten digit recognition using stroke sequences. In this approach, a character instance is represented by a sequence of symbolic strokes, and the representation is obtained by component segmentation and stroke classification. The component segmentation is based on the delta lognormal model of handwriting generation. The symbolic strokes are used for HMM multiple observation training or recognition. A training and recognition experiment has been conducted using the above techniques.

We recently developed a general theory of rapid human movements and applied it to the generation of western handwriting. The goal of this paper is to summarize the key concepts behind the so-called vectorial delta-lognormal model that results from this theory and to show how this model could be used for Chinese character analysis and processing.

Extracting a signature from a check with a patterned background is a thorny problem in image segmentation. Methods based on threshold techniques often necessitate meticulous postprocessing in order to correctly capture the handwritten information. In this study, we tackle the problem of extracting handwritten information by means of an intuitive approach that is close to human visual perception, defining a topological criterion specific to handwritten lines which we call filiformity. This approach was inspired by the existence in the human eye of cells whose specialized task is the extraction of lines. First, we define two topological measures of filiformity for binary objects. Next, we extend these measures to include gray-level images. One of these measures, which is particularly interesting, differentiates the contour lines of objects from the handwritten lines we are trying to isolate. The local value provided by this measure is then processed by global thresholding, taking into account information about the whole image. This processing step ends with a simple fast algorithm. Evaluation of the extraction algorithm carried out on 540 checks with 16 different background patterns demonstrates the robustness of the algorithm, particularly when the background depicts a scene.

This paper focuses on a reading task consisting of the identification of letters in mixed-script handwritten words. This task is performed by humans using extended or limited linguistic context. Their performance rate is to give an upper bound on recognition rates of computer programs designed to recognize handwritten letters in mixed-script writing. Many recognition algorithms are being developed in the research community, and there is a need for establishing ways to compare them. As some effort is on its way to give large test sets with standard formats, we propose an algorithm to determine a test set of reduced size that is appropriate for the task to achieve (the type of texts or words to be recognized). Also, with respect to a particular task, we propose a method for finding an upper limit to the letter recognition rate to aim for.

We propose a method to extract the items of a check by applying three sub processes. First, we eliminate the horizontal lines and some remaining background pixels by a subtraction between a virgin model and a filled specimen of a check, after background elimination on both checks. In the second step, each connected object on the image is described by its minimal surrounding rectangular box. A specific distance is defined for each item. Then the box receives the label of an item if it minimizes its specific distance. In the last step, we propose to reconstruct truncated written lines. The reconstruction is based on a simple geometric hypothesis that a written line does not change its curvature when intersecting with graphic background lines. An evaluation made on 60 different checks shows good results of localization and correct reconstruction when the constant curvature hypothesis is respected.

This paper presents a handwriting generation model based on a kinematic theory of rapid human movements. Handwriting is described as the vectorial sum of formal strokes. These strokes are characterized by a velocity profile that can be described by a delta-lognormal equation. The segmentation algorithm that results from this model allows a comparison of this formal approach with some operational and simpler segmentation methods, in the context of the unicity of the allograph representation.

We exploit the Delta LogNormal theory, a powerful tool for the generation and modeling of rapid movements to generate curvilinear strokes and constituting letters that respect both the dynamics and the appearance of movements made by a human. A theoretical analysis of the effects of the various parameters of the model is carried out: first, to reduce the size of the representation space of the letter models; and second, to select the parameters that constitute the optimal conditions for representing various symbols.

This paper presents a segmentation method that partly mimics the cognitive-behavioral process used by human subjects to recover motor-temporal information from the image of a handwritten word. The approach does not exploit any thinning procedure, but rather a different typology of information is manipulated concerning the curvature of the word contour. Starting from the maximum curvature points roughly corresponding to the beginning of a stroke, the algorithm scans the word, following the natural course of the line and attempts to repeat the same movement as executed by the writer during the generation of the word. At each maximum curvature point, the line is segmented and reconstructed by a smooth interpolation of the most interior points belonging to the line just covered. At the end of the scanning process, a temporal sequence of motor strokes is obtained which plausibly composes the original intended movement.

This paper presents an original method for creating allograph models and recognizing them within cursive handwriting. This method concentrates on the morphological aspect of cursive script recognition. It uses fuzzy-shape grammars to define the morphological characteristics of conventional allographs which can be viewed as basic knowledge for developing a writer independent recognition system. The system uses no linguistic knowledge to output character sequences that possibly correspond to an unknown cursive word input. The recognition method is tested using multi-writer cursive random letter sequences. For a test dataset containing a handwritten cursive text 600 characters in length written by ten different writers, average character recognition rates of 84.4% to 91.6% are obtained, depending on whether only the best character sequence output of the system is considered or if the best of the top 10 is accepted. These results are achieved without any writer-dependent tuning. The same dataset is used to evaluate the performance of human readers. An average recognition rate of 96.0% was reached, using ten different readers, presented with randomized samples of each writer. The worst reader-writer performance was 78.3%. Moreover, results show that system performances are highly correlated with human performances.>

We report the status of the UNIPEN project of data exchange and recognizer benchmarks started two years ago at the initiative of the International Association of Pattern Recognition (Technical Committee 11). The purpose of the project is to propose and implement solutions to the growing need of handwriting samples for online handwriting recognizers used by pen-based computers. Researchers from several companies and universities have agreed on a data format, a platform of data exchange and a protocol for recognizer benchmarks. The online handwriting data of concern may include handprint and cursive from various alphabets (including Latin and Chinese), signatures and pen gestures. These data will be compiled and distributed by the Linguistic Data Consortium. The benchmarks will be arbitrated the US National Institute of Standards and Technologies. We give a brief introduction to the UNIPEN format. We explain the protocol of data exchange and benchmarks.

This paper presents an overview of a dynamic cursive script recognition approach that uses no linguistic constraints. This approach seeks to recognize in cursive script, morphologically and pragmatically coherent sequences of character hypotheses. As performance is compared with the performance of the best available cursive script recognizers-humans-in a reading experiment where linguistic knowledge is useless. The recognition method uses fuzzy-shape grammars to model the morphological characteristics of conventional letters. These models, called allographs, can be viewed as basic (a priori) knowledge for developing a multi-writer recognition system. Character hypotheses are segmented within a cursive word using a parser for these grammars. Character sequences are then constructed from these segmentation hypotheses using local adjacency constraints also modeled by fuzzy-shape grammars. Two experiments are conducted on a test database containing a handwritten cursive test 600 characters in length written by ten different writers. Results show that the system performances are highly correlated with human performance.

This paper describes a system dedicated to online handwritten sentence recognition. The prototype is made up of two basic processors. The first controls the data acquisition, pentip trace segmentation, letter identification. The second aims at identifying and correcting the words candidates by integrating syntactic and lexical information. Sentences are parsed to list the grammatical classes of each incorrect candidates then lexical query searches for words in a lexicon according to grammatical classes. A final decision is made using a string comparison algorithm. Tests of the complete system are reported at the end for a typical writer-dependent application.

This paper is a follow up to an article published in 1989 by R. Plamondon and G. Lorette on the state of the art in automatic signature verification and writer identification. It summarizes the activity from year 1989 to 1993 in automatic signature verification. For this purpose, we report on the different projects dealing with dynamic, static and neural network approaches. In each section, a brief description of the major investigations is given.

This paper reports on the major decisions that have to be made throughout the design phase of an automatic signature verification system. It shows how knowledge about signature control and generation, in terms of a realistic mathematical and psychophysical model, can be helpful in guiding and orienting design choices and related research activities. A system designed according to this methodology is described and the results of a robust forgery test are presented.

The interpretation of handwritten signature images should be closely related to the writer’s identity. The representation and analysis of the handwritten signature is the major challenge in the field of automatic signature verification. A new concept of representation and interpretation of handwritten signature images is advocated. The segmentation process breaks up the signature into a collection of arbitrarily-shaped primitives. In the next step, a local interpretation process serves as a sophisticated template matching, permitting the labeling of all primitives from the test primitive set. This is followed by the global interpretation process, which permits the evaluation of a similarity measure between two structural graphs. Experimental results obtained from a database of 800 handwritten signature images from 20 writers show a performance with a type I error rate of ∈1=1.50%, a type II error rate of ∈2=1.37% and a total error rate ∈t=1.43% in the best strategy proposed using a minimum-distance classifier and two reference signatures. A complete description of this novel automatic handwritten signature verification system is presented in this paper.

A method for comparing thinning algorithms involving a series of experiments with human subjects is presented. Several statistical tests are reported to analyze the preference structure exhibited by the data. It is concluded that humans are coherent in comparing thinning algorithm outputs and that reference skeletons can be useful to facilitate the evaluation of thinning algorithms. An evaluation protocol is then proposed and applied to the design and evaluation of a new thinning algorithm for handwritten character recognition.>

This paper investigates three different methods of comparing preference structures for thinning algorithms. The first method involves a series of experiments with human subjects. The second makes use of neural networks and the third is based on dissimilarities and distance measures that is computer generated. Several statistical tests have been performed to analyze the preference structures exhibited by the data. This study highlights human coherence in comparing skeletons and the novelty of using reference skeletons to facilitate the evaluation of thinning algorithms. None of the automatic approaches provides a useful insight although a measure of information content manifests some consistency. The overall study suggests a systematic protocol involving human coherence to evaluate preprocessing algorithms.

A new algorithm for segmenting continuous handwritten signatures sampled by a digitizer is described. The segmentation points are found using a two-step procedure. The principal step is to construct a function that weights the perceptual importance of every signature point according to its specific neighboring points. The second step points out the various local maxima of this function that correspond to where the signature should be segmented. The method is well illustrated and tested on a number of signatures that require different kinds of segmentation decisions.>

It is generally accepted that highly unstable and easily imitated signatures are among the main causes of deterioration in handwritten signature verification system performance. The authors attempt to estimate, quantitatively and a priori from the coordinates sampled during its execution, the difficulty that could be experienced by a typical imitator in reproducing both visually and dynamically a signature. A functional model of what a typical imitator must do to copy dynamically any signature is derived. A specific difficulty coefficient is then numerically estimated for a given signature. Experimentation geared specifically to signature imitation demonstrates the effectiveness of the model. The ranking of the tested signatures given by the difficulty coefficient is compared to three different sources: the opinions of the imitators themselves, the ones of an expert document examiner, and the ranking given by a specific pattern recognition algorithm. An example of application is also given.>

Defines an operational handwriting model for online syntactic recognition of cursive script. To represent handwriting, it uses characteristic points linked together by segments of uniform curvature. The model acts as a guide for the extraction of attributed primitives used in shape grammars that model allographs and their adjacency rules. The handwriting model is evaluated by human readers in a comparative analysis of original cursive letter sequences versus their reconstructed traces. Also, its performance is measured in terms of mean reconstruction error and data compression rate.>

Describes a model-based segmentation framework for the partitioning of handwriting (handprinted characters, cursive script, signatures). The model accounts for handwriting generation in terms of response patterns that result from the activation by the central nervous system of curvilinear and angular velocity generators, characterized by log-normal impulse responses. In this context, a handwritten trace can be segmented into a hierarchy of well-defined elements: components, strings and curvilinear and angular strokes. One striking conclusion from this approach is that strokes have to be superimposed to generate a smooth handwritten trace and are thus hidden in the trajectory signal. The segmentation algorithm avoids this problem by using an analysis-by-synthesis technique to segment a specific curve.>

This paper presents a real-time constraint-free handprinted character recognition system based on a structural approach. After the preprocessing operation, a chain code is extracted to represent the character. The classification is based on the use of a processor dedicated to string comparison. The average computation time to recognize a character is about 0.07 seconds. During the learning step, the user can define any set of characters or symbols to be recognized by the system. Thus there are no constraints on the handprinting. The experimental tests show a high degree of accuracy (96%) for writer-dependent applications. Comparisons with other system and methods are discussed. We also present a comparison between the processor used in this system and the Wagner and Fischer algorithm. Finally, we describe some applications of the system.

The analysis and synthesis of handwriting are discussed. In the context of a new generation model, a transfer function is proposed to describe the biomechanics of the system involved. This approach allows the extraction of specific movement parameters that reflect the active and passive states of the muscle network (generalized time constant, global gain factor) and the schematic description of the actualized motor program used to generate specific lines (timing and amplitude of the input stimuli). The model is validated with the automatic analysis and synthesis of more than 1500 straight lines written in eight directions by 11 subjects. The parameters extracted during this experiment are analyzed statistically as a function of the direction of movement and of the presence or absence of precue information about these directions. It is shown that asymmetries do exist at different levels in the system. A few directions can be distinguished from the others, but no independent and specific classes of principal directions clearly appear. The most efficient ones correspond to the range of directions of up and down strokes in normal cursive script.>

An online character recognition system based on a comparison of chain codes by a dedicated processor is described. The processing is independent of the size of the characters and the speed of the writing. There is no constraint on the writing, and the system can recognize any character defined by the user. Real-time tests on the system show a high accuracy for user-dependent applications.>

A report is presented on a comparative study of three different signal matching algorithms in the context of signature verification: regional correlation, dynamic time warping, and skeletal tree matching. The algorithm performances are compared in a single experimental protocol over the same database. Algorithm performance is analyzed in terms of verification error rates, execution time, and number and sensitivity of algorithm parameters. Three different script types (normal signatures, handwritten passwords, and initials) and three different signal representation spaces (position, velocity, and acceleration) are considered. Verification errors show that no algorithm consistently outperforms the others in all circumstances.>

A general method is presented for describing and analyzing biomedical handwriting models. Using Laplace's transform theory, a model can be represented in what is called the neural firing-rate domain. Consistent terminology is proposed to facilitate model evaluation and comparison. An overview of previously published models suggests that they could be described using this method, with second- and third-order linear model representation. Fourteen simplified theoretical models are simulated in an experiment designed to study the parameter domain in which handwriting is controlled by the nervous system in order to gain insight into which type of model provides the best reconstruction of natural handwriting. Results show that velocity-controlled models produce the best outputs, with no significant difference between second- and third-order systems. In handwriting, fine motor behavior is thus velocity-controlled. These findings agree with other recent automatic signature verification results and are of interest for a number of applications, from pattern recognition to handwriting education.>

The performance of position, velocity, and acceleration signals for automatic signature verification is compared. Three types of signal comparison algorithms (dynamic time warping, regional correlation and tree matching) are used on a single database. Variance analysis shows that significant differences exist among the three signal representation spaces. Among other things, it is shown that the most discriminant are signals which reflect the vertical activity of the signing process and that the best representation space for a 2-D signature verification system is the velocity domain.>

A preliminary design is given for a general image-understanding system for the extraction of a signature representation. A novel version of an algorithm for centroidal-linkage region-growing with merging, using the statistics of directional data is given. It permits the extraction of textured regions characterized by local uniformity in the orientation of the gradient.>