Nizar Sakr

Prediction-based Haptic Data Reduction and Compression in Tele-Mentoring Systems

In this paper, a novel haptic data reduction and compression technique to reduce haptic data traffic in networked haptic tele-mentoring systems is presented. The suggested method follows a two-step procedure: (1) haptic data packets are not transmitted when they can be predicted within a predefined tolerable error; otherwise, (2) data packets are compressed prior to transmission. The prediction technique relies on the least-squares method. Knowledge from human haptic perception is incorporated into the architecture to assess the perceptual quality of the prediction results. Packet-payload compression is performed using uniform quantization and adaptive Golomb-Rice codes. The preliminary experimental results demonstrate the algorithms effectiveness as great haptic data reduction and compression is achieved, while preserving the overall quality of the tele-mentoring environment.

Robust perception-based data reduction and transmission in telehaptic systems

In this paper, two robust perception-based haptic data reduction and transmission techniques are presented to reduce data traffic in telehaptic systems. A prediction approach that relies on the least-squares method and median filtering is exploited in order to reduce the number of packets transmitted, and efficiently reconstruct unsuccessfully received data samples. Knowledge from human haptic perception is also used and incorporated into the general data reduction architecture. The techniques are initially evaluated in a basic experimental setting in order to validate their performance. Their application in a haptic-enabled telementoring surgery simulation is also demonstrated. The experimental results prove the proposed approachs effectiveness as haptic data packets can be reduced by as much as 96% in normal network conditions and up to 93% in the presence of significant communication delay and packet loss, while preserving the overall quality of the telehaptic environment.

Motion and Force Prediction in Haptic Media

This paper introduces a novel generic method aimed at predicting motion and force information in haptic media. An autoregressive (AR) model is presented for the prediction of both, haptic movement and force. The conditional maximum likelihood technique is utilized in order to accurately estimate the adaptive coefficients of the AR model. Furthermore, the incorporation of concepts from haptic perceptibility, i.e. the just noticeable difference (JND), has been demonstrated to optimize the suggested algorithm, while preserving the immersiveness of the haptic-enabled environment. The proposed technique has also proved to provide accurate prediction results for non-linear haptic movement and force information while simultaneously remaining computationally efficient.

A dynamic fuzzy logic approach to adaptive HVS-based watermarking

A novel image watermarking algorithm is introduced that consists of an adaptive watermarking algorithm based on a human visual system (HVS) model and a dynamic fuzzy inference system (DFIS). This scheme relies on the DFIS to extract the human eye sensitivity knowledge using the HVS model. The DFIS and the HVS combined are used to adjust and select the appropriate watermark length as well as the watermark strength for each pixel in an image. The main goal of the algorithm is to provide a more robust and imperceptible watermark. The aforementioned has been implemented and tested under various attacks including image compression, cropping, additive Gaussian noise distortion, scaling, low-pass filtering, as well as collusion attacks. The results achieved demonstrate that the watermark can survive these attacks while remaining imperceptible. The primary contribution of this adaptive watermarking approach is to further enhance spread-spectrum based watermarking schemes by introducing a DFIS model that encompasses a dynamic membership function engine to accurately approximate the relationship found between all properties of the HVS.

Human Perception-Based Data Reduction for Haptic Communication in Six-DoF Telepresence Systems

In this paper, a human perception-based data reduction method is suggested to reduce the number of packets transmitted in 6-degrees-of-freedom (DoF) telehaptic systems; specifically in haptic-enabled telepresence. The algorithm relies on knowledge from human haptic perception in order to reduce the number of packets transmitted without compromising transparency. Several distance metrics are also discussed to best examine the acuity of human perception in detecting haptic distortion when data reduction is performed in 6-DoF settings. A validation of the proposed haptic data reduction technique is performed under normal network conditions as well as in the presence of network-induced time delay and packet loss. Statistical significance tests (using Friedmans nonparametric ANOVA, and Wilcoxon signed-rank tests) were carried out to determine the appropriate multivariate human haptic perceptual thresholds (force, torque, orientation, etc.) required to minimize the number of packets transmitted while preserving the immersiveness of the 6-DoF telehaptic environment. It was observed that the suggested algorithm can significantly reduce haptic data traffic with little or no influence on the quality of haptic-enabled telepresence interaction.

Towards an Architecture for the Compression of Haptic Media

In this paper, we present I-CHAM, an architecture for the intelligent compression of haptic media. The I-CHAM architecture enables both, lossy and lossless compression modes and it encompasses several modules that, together, attempt to intelligently and efficiently compress a haptic media file. Furthermore, the proposed architecture is designed in such a manner to be haptic-application and device independent. To the best of our knowledge, this paper is the first in the literature to propose an architecture for the compression of haptic media files. The experimental results demonstrate that I-CHAM can achieve great compression ratios in both, lossless and lossy modes, while preserving the immersiveness of the haptic-based environment.

Feature selection and classification in genetic programming: Application to haptic-based biometric data

In this paper, a study is conducted in order to explore the use of genetic programming, in particular gene expression programming (GEP), in finding analytic functions that can behave as classifiers in high-dimensional haptic feature spaces. More importantly, the determined explicit functions are used in discovering minimal knowledge-preserving subsets of features from very high dimensional haptic datasets, thus acting as general dimensionality reducers. This approach is applied to the haptic-based biometrics problem; namely, in user identity verification. GEP models are initially generated using the original haptic biometric datatset, which is imbalanced in terms of the number of representative instances of each class. This procedure was repeated while considering an under-sampled (balanced) version of the datasets. The results demonstrated that for all datasets, whether imbalanced or under-sampled, a certain number (on average) of perfect classification models were determined. In addition, using GEP, great feature reduction was achieved as the generated analytic functions (classifiers) exploited only a small fraction of the available features.

A Perceptual Quality Metric for Haptic Signals

In this paper, an objective perceptual quality metric for haptic signals is introduced. A haptic perceptually weighted peak signal-to-noise ratio (HPWNPSNR) is derived, that enhances the standard PSNR measure while incorporating haptic perceptual knowledge acquired from existing psychophysical models. Two techniques are introduced that define the haptic perceptual weighting strategy. The first relies on a basic mathematical formulation, whereas the alternative technique makes use of fuzzy logic principles. Preliminary results are demonstrated to validate the proposed metrics utility and performance.

Feature selection in haptic-based handwritten signatures using rough sets

This paper explores the use of rough set theory for feature selection in high dimensional haptic-based handwritten signatures (exploited for user identification). Two rough set-based methods for feature selection are analyzed, the first is a greedy approach while the second relies on genetic algorithms to find minimal subsets of attributes. Also, to further reduce the haptic feature space while maximizing user identification accuracy, a method is proposed where feature vectors are subsampled prior to the feature selection procedure. Rough setgenerated minimal subsets are initially exploited to determine the importance of different haptic data types (e.g. force, position, torque and orientation) in discriminating between different users. In addition, a comparison between rough set-based methods and classical machine learning techniques in the selection of minimal information-preserving subsets of features in high dimensional haptic datasets, is provided. The criteria for comparison are the length of the selected subsets of features and their corresponding discrimination power. Support Vector Machine classifiers are used to evaluate the accuracy of the selected minimal feature vectors. The results demonstrated that the combination of rough set and genetic algorithm techniques can outperform well-established machine learning methods in the selection of minimal subsets of features present in haptic-based handwritten signatures.

Exploring the underlying structure of haptic-based handwritten signatures using visual data mining techniques

In this paper, multidimensional and time-varying haptic-based handwritten signatures are analyzed within a visual data mining paradigm while relying on unsupervised construction of virtual reality spaces using classical optimization and genetic programming. Specifically, the suggested approaches make use of nonlinear transformations to map a high dimensional feature space into another space of smaller dimension while minimizing some error measure of information loss. A comparison between genetic programming and classical optimization techniques in the construction of visual spaces using large haptic datasets, is provided. In addition, different distance functions (used in the nonlinear mapping procedure between the original and visual spaces) are examined to explore whether the choice of measure affects the representation accuracy of the computed visual spaces. Furthermore, different classifiers (Support Vector Machines (SVM), k-nearest neighbors (k-NN), and Nai¿ve Bayes) are exploited in order to evaluate the potential discrimination power of the generated attributes. The results show that the relationships between the haptic data objects and their classes can be appreciated in most of the obtained spaces regardless of the mapping error. Also, spaces computed using classical optimization resulted in lower mapping errors and better discrimination power than genetic programming, but the later provides explicit equations relating the original and the new spaces.