Andreja Samčović

H.264/AVC video compressed traces: multifractal and fractal analysis

Publicly available long video traces encoded according to H.264/AVC were analyzed from the fractal and multifractal points of view. It was shown that such video traces, as compressed videos (H.261, H.263, and MPEG-4 Version 2) exhibit inherent long-range dependency, that is, fractal, property. Moreover they have high bit rate variability, particularly at higher compression ratios. Such signals may be better characterized by multifractal (MF) analysis, since this approach describes both local and global features of the process. From multifractal spectra of the frame size video traces it was shown that higher compression ratio produces broader and less regular MF spectra, indicating to higher MF nature and the existence of additive components in video traces. Considering individual frames (I, P, and B) and their MF spectra one can approve additive nature of compressed video and the particular influence of these frames to a whole MF spectrum. Since compressed video occupies a main part of transmission bandwidth, results obtained from MF analysis of compressed video may contribute to more accurate modeling of modern teletraffic. Moreover, by appropriate choice of the method for estimating MF quantities, an inverse MF analysis is possible, that means, from a once derived MF spectrum of observed signal it is possible to recognize and extract parts of the signal which are characterized by particular values of multifractal parameters. Intensive simulations and results obtained confirm the applicability and efficiency of MF analysis of compressed video.

Face Detection Approach in Neural Network Based Method for Video Surveillance

Neural networks are adaptive information processing systems that offer attractive solutions for video surveillance. This application aims at identifying particular patterns. Also, MPEG-4 standard profiling strategy in facial animation guarantees that the standard can provide adequate solutions for video surveillance. The main goal of this presentation is to provide face detection for video surveillance using neural network based method. After providing the corresponding architecture for face detection, the emphasize is on the detector which is trained with multilayer back propagation neural networks. Three different face representations are taken into account, i.e. pixel representation, partial profile representation and eigenface representation. Based on this, three independent sub-detectors are generated. The detection rates are measured. The circle at about 94% indicates the position where the neural network achieves the optimal performance

Exploring college students' generational differences in Facebook usage

In recent years the usage of social networks among students has become the subject of numerous studies. In this study we were interested in differences in patterns of online behaviour and the use of Facebook (FB) among the second and third generation of students. The results showed that 66% of the second year students had a FB account, while in the third generation all of the respondents had a FB account. One third of the second year students avoid any involvement in FB. Lack of interest was one of the main reasons the respondents indicated for not being engaged in FB, as well as their unwillingness to put personal information online. Currently, most of them are satisfied using FB, and do not plan to turn off their profile. The newer generations of students use FB mainly for overcoming boredom. Based on the results of the study, it can be concluded that the time spent online increases, as well as the number of FB users, and also the number of virtual friends increases in order to exchange information. We concluded that the number of users of Facebook increases among older students in comparison to younger one.We found that increasing the number of friends on Facebook is more pronounced among older students.Membership in groups is more popular within older students.Most students do not consider that they are addicted to Facebook.

Vision-based interaction as an input of serious game for motor rehabilitation

In long-term rehabilitation process is habitual user demotivation, because it usually consists in repetitive and intensive activities which become boring after hundreds of sessions. Research studies showed that serious games help to motivate users in rehabilitation process. Motor rehabilitation consists of body movements that can be captured and patients can have difficulties in holding physical devices. For this reason, therapy studies have included vision input devices and proposed what features were desirable for rehabilitation serious games. In this paper, we present the benefits of vision-based interaction for rehabilitation and a case of success, a vision-based interactive system for balance and postural of adults with cerebral palsy.

Film archives as a future enhancement of digital cinema technology

Starting from the point of view that digital cinema offers an enhanced viewing experience for audiences, content flexibility, and distribution cost savings, this work deals with film archives as a future enhancement of digital cinema technology. Since the transition from physical media to electronic one represents a paradigm shift in video compression techniques and applications, archival requirements together with the system architecture are analyzed. Features extractions for shot cut detection in the film material are presented, too. Some simulation results for JPEG 2000 as well as H.264 standards including PSNR versus bit rate for the 4K resolution content and the low resolution 512 sequence conclude the paper.

Analysis of Coding Gain and Optimal Bit Allocation in Motion-Compensated Video Compression

Analysis of Coding Gain and Optimal Bit Allocation in Motion-Compensated Video Compression This paper describes mathematical frameworks on temporal and spatial predictive processing in the motion-compensated video compression. Firstly, the coding gain over intra coding is derived, regarding the bit allocation algorithm and Lagrange multiplier method. The optimal ordering of three different picture types (I, P and B pictures) is clarified according to image source characteristics. Secondly, a novel framework with the block-based multihypothesis motion-compensated optimal coding gain and bit allocation are derived in a closed-form expression.

UPPER BOUND ON SNR GAIN IN WAVELET TRANSFORM PREDICTIVE-ENTROPY IMAGE CODING

Using the rate-distortion theory approach and assuming the Laplacian probability density function of the quantizer input signal, we propose a function which gives the signal-to-quantizing-noise ratio (SNR) gain in discrete wavelet transform predictive-entropy coding over fullband predictive image coding. The upper bound on SNR gain is determined as a function of the subband number only. The practical SNR performances of realizable wavelet still image coders are compared with their theoretical bounds. The computer simulation results of wavelet based predictive coded test images show that the SNR gain grows faster with the subband number increment than its theoretical upper bounds do.

Subjective image quality estimation in some subband coding techniques

This paper describes some subband coding techniques from the users point of view, concerning subjective image quality estimation. The characteristics of a differential pulse code modulation, as a method to encode subbands, are also described. Subband coding with discrete cosine transform is presented, too. Finally, in order to estimate the image quality, some experimental results are demonstrated. Mean opinion score quality scale with five degrees is recommended.

System Usability Scale Implementation for Interfaces on Mobile Touch Screen Devices Assessment

The indisputable fact is that touch screen mobile devices flooded the marketplace and that they have become an essential element in the enormous number of fields and applications. But a huge presence, as well as daily use, does not necessarily unambiguously reflect their optimal usability. Therefore, it is essential to adopt appropriate research methodologies that can evaluate the usability of mobile device interfaces. The main focus of the study was to examine a practical tool for usability assessment in a mobile work context which has to be simple and above all, proved to give a single score as its outcome. To achieve this goal System Usability Scale (SUS) was applied to test the Tablet PC usability. This study contributed with the proposal of a framework for usability testing which could help, both practitioners and researchers to make decisions how to improve the usability of interfaces on touch screen mobile devices.

Application of artificial neural networks for slope stability analysis in geotechnical practice

In present paper, authors develop a model for estimation of earth slope stability based on the artificial neural networks. For this purpose, authors engage multi-layer feed-forward network with Levenberg-Marquardt learning algorithm and 14 hidden nodes, using existing experimental data, and the results of traditional limit equilibrium analyzes of 57 different cases according to the predefined experimental plan. The results obtained indicate high level of statistical reliability (R=0.95 and MSE=0.0035 for testing set of scaled values) and similar estimation accuracy as the existing mathematical expression for calculation of slope safety factor.