Paraskevi K. Tzouveli

Estimation of behavioral user state based on eye gaze and head pose--application in an e-learning environment

Most e-learning environments which utilize user feedback or profiles, collect such information based on questionnaires, resulting very often in incomplete answers, and sometimes deliberate misleading input. In this work, we present a mechanism which compiles feedback related to the behavioral state of the user (e.g. level of interest) in the context of reading an electronic document; this is achieved using a non-intrusive scheme, which uses a simple web camera to detect and track the head, eye and hand movements and provides an estimation of the level of interest and engagement with the use of a neuro-fuzzy network initialized from evidence from the idea of Theory of Mind and trained from expert-annotated data. The user does not need to interact with the proposed system, and can act as if she was not monitored at all. The proposed scheme is tested in an e-learning environment, in order to adapt the presentation of the content to the user profile and current behavioral state. Experiments show that the proposed system detects reading- and attention-related user states very effectively, in a testbed where children’s reading performance is tracked.

Adaptive Reading Assistance for the Inclusion of Students with Dyslexia: The AGENT-DYSL Approach

Dyslexia is a major barrier to success in education and later on the job as reading skills are fundamental for personal competence development. Children with dyslexia have special learning needs (e.g., more teacher support), which currently only specialized institutions can provide. However, this takes children out of their peer group and causes social problems. On the other side, there is general-purpose reading support software, which are not geared towards children with dyslexia as they lack personalization. AGENT-DYSL brings together speech and image recognition as well as semantic technologies to build a truly adaptive reading support system for children with dyslexia.

Human face watermarking based on Zernike moments

A novel human face watermarking scheme is proposed in this paper, providing copyright protection of semantic content. To achieve this goal, skin detection is initially performed using a skin filter, which relies on color information and then, face extraction is achieved using a combination of a morphological filter and a human face template. An invariant watermark is then designed and tested against attacks using invariant Zernike moments. The proposed algorithm has the advantages of being robust, computationally efficient and overheads transmitted to the decoder side are very low. The performance of the proposed human face watermarking system is tested under various signal distortions such as JPEG lossy compression, blurring, filtering and cropping. Experimental results on real life images indicate the efficiency and robustness of the proposed scheme

Semantic Classification of Byzantine Icons

This system uses fuzzy description logics and patterns to automatically determine the sacred figure depicted in an icon.

Non-Verbal Feedback on User Interest Based on Gaze Direction and Head Pose

Farsi language is one of the dominant languages in middle-east. A lot of work has been done on Farsi retrieval systems. Local context analysis is a query expansion method to improve retrieval performance. In this paper we have tried to tune LCA for Farsi language. We used Hamshahri collection and 60 queries to tune three parameters in LCA method which are number of concepts used for query expansion, number of initially retrieved documents for local feedback and number of passages for concept discovery and weighting. The results reveal that there is a possible optimization point when 20 concepts are used; however, increasing the other two parameters which are number retrieved documents and number of passages used for local feedback almost always yields better results.

E-learning and intelligent content adaptation: an integrated approach

It is a common fact that modern e-learning schemes lack educational content representation and user personalisation. In this framework, automated extraction of user profiles, to be used in an e-learning content offering system, forms an interesting and important problem. In this approach we present the design and implementation of such a profile-based system, by which content is matched to its environmental context, so that it can be adapted to its users needs and capabilities. Current effort extends previous work on profile extraction via clustering techniques and on integrated e-learning systems. It relies on the fundamental IEEE e-learning model, suitably adapted to reflect and focus on profiling aspects of the system.

Video object watermarking based on moments

A robust video object based watermarking scheme, based on Zernike and Hu moments, is proposed in this paper. Firstly, a human video object detector is applied to the initial image. Zernike and the Hu moments of each human video object are estimated and an invariant function for watermarking is incorporated. Then, the watermark is generated modifying the moment values of each human video object. In the detection scheme, a neural network classifier is initially used in order to extract possible watermarked human video objects from each received input image. Then, a watermark detection procedure is applied for video object authentication. A full experiment confirms the promising performance of the proposed scheme. Furthermore, the performances of the two types of moments are extensively investigated under several attacks, verifying the robustness of Zernike moments comparing to Hu moments.

AUTOMATIC FACE REGION WATERMARKING USING QUALIFIED SIGNIFICANT WAVELET TREES

In this paper, a fully automatic scheme for hiding digital watermarks into face regions is proposed. To achieve this goal, an adaptive two-dimensional Gaussian model of skin color distribution is initially used in order to detect face regions within the initial image. Next each face region is decomposed into three levels with ten subbands, using the Discrete Wavelet Transform (DWT) and three pairs of subbands are formed (HL3, HL2), (LH3, LH2) and (HH3, HH2). Afterwards Qualified Significant Wavelet Trees (QSWTs), which are derived from the Embedded Zerotree Wavelet (EZW) algorithm and they are high-energy coefficient paths, are estimated for a pair of subbands. Finally visually recognizable watermark patterns are redundantly embedded to the coefficients of the highest energy QSWTs and the IDWT is applied to provide the watermarked face area. Performance of the proposed face region watermarking system is tested under various signal distortions such as JPEG lossy compression, sharpening and blurring. Experimental results on real life images indicate the efficiency and robustness of the proposed scheme.

Affine invariant curve matching using normalization and curvature scale-space

In this paper, an affine invariant curve matching method using curvature scale-space and normalization is proposed. Prior to curve matching, curve normalization with respect to affine transformations is applied, allowing a lossless affine invariant curve representation. The maxima points of the curvature scale-space (CSS) image are then used to represent the normalized curve, while retaining the local properties of the curve. The matching algorithm that follows, matches the maxima sets of CSS images and the resulting matching cost provides a measure of similarity. The methodpsilas performance and robustness is evaluated through a variety of curves and affine transformations, obtaining precise shape similarity and retrieval.

Human video object watermarking based on Hu moments

A novel video object based watermarking scheme is proposed in this paper, providing copyright protection of the semantic content. To achieve this goal, an adaptive two-dimensional Gaussian model of skin color distribution is initially used in order to detect face and body regions within the initial image. An invariant watermark is then designed and tested against attacks using invariant Hu moments. The proposed algorithms have the advantages of being robust, computationally efficient, and overheads transmitted to the decoder side are very low. Performance of the proposed object based watermarking system is tested under various signal distortions such as JPEG lossy compression, blurring, filtering and cropping, Experimental results on real life images indicate the efficiency and robustness of the proposed scheme.