SIGMAP 2012 Abstracts

Area 1 - Multimedia and Communications
Short Papers
Area 2 - Multimedia Signal Processing
Full Papers Short Papers
Area 3 - Multimedia Systems and Applications
Full Papers Short Papers

Area 1 - Multimedia and Communications

Short Papers
Paper Nr: 18
Title:

Adaptive Rate Control Scheme for Improving Quality of Multimedia in Broadband Wireless Networks
Authors:

Dooyeol Yoon, Dongchil Kim and Kwangsue Chung
Abstract: In order to improve quality of streaming services in broadband wireless networks, many researches are in progress. However, existing schemes do not guarantee a user perceived quality, because most of these schemes do not consider both wireless channel states and video characteristics. To cope with these problems, this paper proposes a NB-RC (Network and Buffer-aware Rate Control) scheme. The proposed scheme adjusts the video transmission rate according to the wireless channel states. It also controls the video quality based on buffer occupancy of clients. Through the simulation results, we prove that our scheme improves the media quality.
Download
Paper Nr: 57
Title:

Time-frequency Filtering of Gaussian and Impulse Noise for Spread Spectrum Power Line Communication
Authors:

Gaoyong Luo
Abstract: The affluence of impulse noise is one of the challenging problems of the power line communication (PLC) as a communication channel. However, current methods for impulse noise reduction are either not effective or requiring heavy computing for detecting impulse noise accurately. This paper presents a time-frequency filter design method for impulse and Gaussian noise mitigation by a reliable noise detector in the wavelet domain with local variance analysis. The filtering is applied only to the detected noisy samples with others unchanged in an effort to reduce the noise level by adapting its operation in accordance with variance characteristics. The received corrupted signal from spread spectrum system is decomposed into time-frequency domain by fast implementation of lifting wavelet transform for real-time filtering of mixed Gaussian and impulse noise. Experimental results demonstrate that the proposed method can significantly reduce impulse noise and improve bit error rate (BER) without introducing distortion, leading to better quality of service.
Download
Paper Nr: 75
Title:

Spaxels, Pixels in Space - A Novel Mode of Spatial Display
Authors:

Horst Hörtner, Matthew Gardiner, Roland Haring, Christopher Lindinger and Florian Berger
Abstract: We introduce a novel visual display paradigm through the use of controllable moving visible objects in physical space. Spaxels is a conjugation of "space" and "pixels". It takes the notion of a pixel, and frees it from the confines of a static two-dimensional matrix of a screen or projection surface to move three dimensionally in space. Spaxels extend the notion of Voxels, volumetric pixels, in that a Spaxel can physically move, as well as transition in colour and shape. Our current applied research is based on the control of a swarm of unmanned aerial vehicles equipped with RGB lighting and a positioning system that can be coordinated in three dimensions to create a morphing floating display. This paper introduces Spaxels as a novel concept and paradigm as a new kind of spatial display.
Download
Paper Nr: 11
Title:

MPEG-4/AVC versus MPEG-2 in IPTV
Authors:

Stefan Paulsen, Tadeus Uhl and Krzysztof Nowicki
Abstract: This paper is essentially a treatment of the theoretical and practical aspects of the new IPTV service. The central part of the paper constitutes a detailed presentation of analysis scenarios and results, and addresses the following issues in particular: What influence does the encoding rate have of on QoE values? What effect does the most obtrusive impairment factor in a network, i.e. packet loss, have on QoE in IPTV? Is the MPEG-2 Transport Stream suitable for encapsulation and transport of MPEG-4/AVC content? Are there alternatives to the ISO/IEC 13818-1 Transport Stream? If so, how do they affect quality of service (QoE)?
Download
Paper Nr: 12
Title:

Design of Short Irregular LDPC Codes for a Markov-modulated Gaussian Channel
Authors:

W. Proß, M. Otesteanu and F. Quint
Abstract: This paper deals with the design of short irregular Low-Density Parity-Check (LDPC) codes. An optimization method for the underlying symbol-node degree-distribution (SNDD) of an irregular LDPC code is introduced, which is based on the Downhill-Simplex (DHS) algorithm. In order to compare our method with the optimization described in (Hu et al., 2005), which is based on a simplified version of the DHS algorithm, we first designed a rate 0.5 irregular LDPC code of length n = 504 for an Additive White Gaussian Noise Channel (AWGNC). The proposed optimization method was then used to design an irregular LDPC code for a Markov-modulated Gaussian Channel (MMGC). The decoding performance of the resulting LDPC code is then compared to the design based on the Density-Evolution (DE) method.
Download
Paper Nr: 40
Title:

Enhancing Photoware in the Social Networks Environment
Authors:

Ombretta Gaggi
Abstract: This paper presents SMIL PhotoShow, an authoring tool for photo books, which allows the creation of enhanced multimedia presentations, enriched with audio (music or spoken) comments, transition effects and animations in a very simple way. Our tool allows to create a digital counterpart of a printed photo book with the aim to bridge the gap between digital web albums and printed photo books. Since authoring a photo book is a time consuming activity, SMIL PhotoShow provides the users also the possibility to create, with only three clicks, an engaging slideshows with transition effects and background music.
Download

Area 2 - Multimedia Signal Processing

Full Papers
Paper Nr: 44
Title:

Diffusion Tracking Algorithm for Image Segmentation
Authors:

Lassi Korhonen and Keijo Ruotsalainen
Abstract: Different clustering algorithms are widely used for image segmentation. In recent years, spectral clustering has risen among the most popular methods in the field of clustering and has also been included in many image segmentation algorithms. However, the classical spectral clustering algorithms have their own weaknesses, which affect directly to the accuracy of the data partitioning. In this paper, a novel clustering method, that overcomes some of these problems, is proposed. The method is based on tracking the time evolution of the connections between data points inside each cluster separately. This enables the algorithm proposed to perform well also in the case when the clusters have different inner geometries. In addition to that, this method suits especially well for image segmentation using the color and texture information extracted from small regions called patches around each pixel. The nature of the algorithm allows to join the segmentation results reliably from different sources. The color image segmentation algorithm proposed in this paper takes advantage from this property by segmenting the same image several times with different pixel alignments and joining the results. The performance of our algorithm can be seen from the results provided at the end of this paper.
Download
Paper Nr: 58
Title:

Optimisation of Smoothing Parameter of Diffeomorphism Kernel Estimate for Bounded Random Data
Authors:

Molka Troudi and Faouzi Ghorbel
Abstract: The Diffeomorphism Kernel Density Estimator (DKDE) requires the estimation of an optimal value of the bandwidth to ensure a reliable pdf estimation of bounded distributions. In this paper, we suggest to approach the optimal bandwidth value by adapting Plug-in algorithm to DKDE estimator. We will show that the pro-posal method allows better density estimation in the MISE sense. Otherwise, the Gibbs phenomenon com-pletely disappears. These results are illustrated by some bounded and semi bounded distributions simulations.
Download
Paper Nr: 60
Title:

Development of Computer Algorithms to Control a Weelchair through the Movement of the Head by Artificial Vision
Authors:

Ricardo Fuentes Covarrubias, Andrés Gerardo Fuentes Covarrubias, Cristina Conde Vilda, Isaac Martin de Diego and Enrique Cabello
Abstract: The Purpose of this project is the control of motion and direction in real time of a wheel chair, using machine vision algorithms. The main goal of this project is the signal acquisition from the video camera and collision sensors for post processing in the C# algorithms and later obtaining motor control in the traction mechanism of the wheelchair. The C# algorithm has several tasks. The first is to obtain the real time image from web cam and later processing for the identification of the direction of movement of the human face. The second is to calculate the speed of the movement for generation of the PWM output for motor movement. This information output using the RS232C driver to a microcontroller card attached to a motor control box in the wheel chair mechanism. The final task is to obtain the collision sensor status for security implementation, all in real time. The main reason for development of an implementation of this solution is the use of open source software tools for a more stable platform in the base system due to the characteristics of the end use of the system. The end user of the system will be quadriplegic people.
Download
Short Papers
Paper Nr: 13
Title:

Fast Algorithm of Short-time DCT for Low Resolution Signal Processing
Authors:

Vitaly Kober
Abstract: A fast algorithm for computing the discrete cosine transform (DCT) in a window running on a signal with a step higher than one is proposed. The algorithm is based on a second-order recursive relation between DCT spectra computed in windows which are equally spaced with a given distance. The computational complexity of the proposed algorithm is compared with that of common fast and running DCT algorithms. A fast inverse DCT transform is also presented.
Download
Paper Nr: 38
Title:

Sphere Decoding Complexity Reduction using an Adaptive SD-OSIC Algorithm
Authors:

Bora Kim, Sangmi Moon, Saransh Malik, Cheolhong Kim and Intae Hwang
Abstract: Sphere decoding is a technique able to achieve the optimal performance of the maximum likelihood decoder, but its high and variable complexity can make the practical implementation infeasible. In this paper, we present an adaptive system, called adaptive SD-OSIC, as a way of reducing the decoding complexity while maintaining the error performance of conventional sphere decoding.
Download
Paper Nr: 46
Title:

Finding a Tradeoff between Compression and Loss in Motion Compensated Video Coding
Authors:

Thomas Guthier, Adrian Sosic, Volker Willert and Julian Eggert
Abstract: In video coding, affine motion models combined with a quadtree decomposition have often been suggested as an extension to the mostly used translational models combined with a blockwise decomposition. What is missing so far is a thorough analysis to judge the tradeoff between using more complex motion models or more elaborate decomposition methods in terms of data compression and information loss. In this paper, we compare different polynomial motion models with a quadtree decomposition concerning motion model complexity and granularity of decomposition. We provide a statistical evaluation based on optical flow databases to quantitatively find a tradeoff between bitrate and reconstruction error.
Download
Paper Nr: 54
Title:

New Two-step Motion Estimation using Adjustable Partial Distortion Search - Advanced Selected Search Point and Early Termination for Two Step Motion Search
Authors:

Yonghoon Kim and Jechang Jeong
Abstract: In this paper, we proposed an advanced two-step motion estimation using adjustable partial distortion for fast motion estimation. We improve the two-step search by using relationship between neighboring and current block. The proposed algorithm is 187 times faster than FS and 2.7 times faster than TS-EPDS without negligible PSNR degradation. Therefore, it is suitable for real-time video implementation.
Download
Paper Nr: 55
Title:

Video Foreground/Background Segmentation using Spatially Distributed Model and Edge-based Shadow Cancellation
Authors:

Shian-De Tsai, Jin-Jang Leou and Han-Hui Hsiao
Abstract: Video foreground/background segmentation is to extract relevant objects (the foreground) from the background of a video sequence, which is an important step in many computer vision applications. In this study, the spatially distributed model is built by a splitting process using Gaussian probability distribution functions in spatial and color spaces. Then, edge-based shadow cancellation is employed to obtain more robust segmentation results. The proposed approach can well handle illumination variations, shadow effect, and dynamic scenes in video sequences. Based on experimental results obtained in this study, as compared with two comparison approaches, the proposed approach provides the better video segmentation results.
Download
Paper Nr: 61
Title:

Raw Camera Image Demosaicing using Finite Impulse Response Filtering on Commodity GPU Hardware using CUDA
Authors:

Patrik Goorts, Sammy Rogmans and Philippe Bekaert
Abstract: In this paper, we investigate demosaicing of raw camera images on parallel architectures using CUDA. To generate high-quality results, we use the method of Malvar et al., which incorporates the gradient for edgesensing demosaicing. The method can be implemented as a collection of finite impulse response filters, which can easily be mapped to a parallel architecture. We investigated different trade-offs between memory operations and processor occupation to acquire maximum performance, and found a clear difference in optimization principles between different GPU architecture designs. We show that trade-offs are still important and not straightforward when using systems with massive fast processors and slower memory.
Download
Paper Nr: 62
Title:

A Dataflow Description of ACC-JPEG Codec
Authors:

Khaled Jerbi, Tarek Ouni and Mohamed Abid
Abstract: Video codec standards evolution raises two major problems. The first one is the design complexity which makes very difficult the video coders implementation. The second is the computing capability demanding which requires complex and advanced architectures. To decline the first problem, MPEG normalized the Reconfigurable Video Coding (RVC) standard which allows the reutilization of some generic image processing modules for advanced video coders. However, the second problem still remains unsolved. Actually, technology development becomes incapable to answer the current standards algorithmic increasing complexity. In this paper, we propose an efficient solution for the two problems by applying the RVC methodology and its associated tools on a new video coding model called Accordion based video coding. The main advantage of this video coding model consists in its capacity of providing high compression efficiency with low complexity which is able to resolve the second video coding problem.
Download
Paper Nr: 8
Title:

Complexity Analysis of Video Frames by Corresponding Audio Features
Authors:

SeungHo Shin and TaeYong Kim
Abstract: In this paper, we propose a method to estimate the video complexity by using audio features based on human synesthesia factors. By analyzing the features of audio segments related to video frames, we initially estimate the complexity of the video frames and can improve the performance of video compression. The effectiveness of proposed method is verified by applying it to an actual H.264/AVC Rate-Control.
Download
Paper Nr: 35
Title:

Performance Analysis of New SNR Estimation Methodology based on Preamble Approach
Authors:

Sangmi Moon, Saransh Malik, Bora Kim, Cheolhong Kim and Intae Hwang
Abstract: The latest wireless communication systems focus on developing MIMO-OFDM systems that allow the transmission of very high data rates in fading environments. We can optimize these systems even further by setting the modulation and coding adaptively according to the channel conditions, and by using sub-carrier frequency and power allocation techniques. The overall system performance depends on the accuracy and delay of the channel state information (CSI). In this paper, we propose a signal-to-noise ratio (SNR) estimation algorithm based on preamble transmission. Through simulations of several channel environments, we prove that our proposed algorithm is more accurate than conventional algorithms.
Download
Paper Nr: 36
Title:

Novel Channel Estimation Algorithm using Various Filter Design in LTE-Advanced System
Authors:

Saransh Malik, Sangmi Moon, Bora Kim, Cheolhong Kim and Intae Hwang
Abstract: Channel estimation is a major issue in communication system. In this paper, we propose a new idea for channel estimation that uses a Kalman Filter (KF) approach to predict the channel in OFDM symbols with pilot subcarriers where channel affected is by high doppler spread. We design the algorithm considering the lattice-type arrangement of pilot subcarriers in a LTE-Advanced system from 3GPP. In further advancement, we use the filtering of channel impulse response and application of a Wiener Filter for the estimation of the channel frequency response in the rest of the subcarriers.
Download
Paper Nr: 45
Title:

Optimal Multidimensional Signal Processing in Wireless Sensor Networks
Authors:

Anatoli Torokhti and Stan Miklavcic
Abstract: Wireless sensor networks involve a set of spatially distributed sensors and a fusion center. Three methods for finding models of the sensors and the fusion center are proposed.
Download
Paper Nr: 51
Title:

Search Range Adjustment and Motion Vector Prediction for Fast Motion Estimation - Using Neighbouring Motion Vectors and Distortions for Adjustment of a Search Range and a Starting Point
Authors:

Do-Kyung Lee and Je-Chang Jeong
Abstract: The block matching algorithm has been widely used for motion estimation, but it suffers from heavy computational complexity. Many researchers in video processing area have proposed fast motion estimation algorithms by adopting various ways to reduce its complexity. In this paper, we introduce a conventional method controlling a search range and defining a new starting point, and also discuss our proposed method which supplements previous work by using neighbouring block’s motion vectors and distortions. Consequently, we obtained remarkable performance about 80 times faster than full search and 2.6 times faster than conventional algorithm with small video quality improvement in terms of PSNR. Therefore, the proposed method can be applied for real-time video processing applications.
Download
Paper Nr: 65
Title:

Friction Sources Characterization for Fricative Consonants of Arabic
Authors:

Fazia Karaoui and Amar Djéradi
Abstract: The objective of this work is the acoustic characterization of the friction source for Arabic voiced consonants [v], [z], [ʒ], [h], [ς] and unvoiced ones [s], [f], [∫], [ħ], employing vocal tract transfer function obtained from a direct measurement by the Pseudo Random Excitation of the human vocal tract and the signal spectrum radiated at the lips. Assuming the separability of the source of the vocal tract considered as a linear filter, the sources spectrum is obtained by the ratio of the output signal spectrum of the vocal tract transfer function. The results are derived from data produced by two female and two male subjects.
Download

Area 3 - Multimedia Systems and Applications

Full Papers
Paper Nr: 59
Title:

Keywords-based Automatic Multimedia Authoring in the Cloud
Authors:

Abdelkader Outtagarts, Sylvain Squedin and Olivier Martimot
Abstract: In this paper, we describe our work on automatic multimedia authoring or editing. Our approach is based on keyword extracted from the audio embedded in videos. A model approach of mashup based on keywords is proposed. A video editing testbed has been designed and implemented. Using speech2text keywords generator component, the audio files uploaded in the video editing testbed are transcribed and analyzed in order to extract keywords. The keywords are used to edit automatically videos in order to produce mashups.
Download
Short Papers
Paper Nr: 33
Title:

The Effect of Multi-media Contents in Reducing Sensible Temperature
Authors:

Shuhei Yamamoto, Akira Tomono and Hajime Katsuyama
Abstract: In this paper, the effect of multi-media contents such as visual images, scent, and their combinations on sensible temperature is investigated. For this purpose, a new definition of sensible temperature which takes into account the effect of visual images and scent is proposed. Using this definition, the effectiveness of multi-media contents in reducing sensible temperature was quantitatively measured. It turned out that visual images with lemon aroma is more effective in reducing sensible temperature than visual images alone.
Download
Paper Nr: 34
Title:

Image Presentation with Smell for Digital Signage and the Effect on Eye Catching
Authors:

Keisuke Tomono, Hajime Katsuyama, Shuhei Yamamoto and Akira Tomono
Abstract: This paper describes the effect on eye catching by digital signage that releases smell from screen and the method of smell presentation to human olfactory receptor. The effect on eye catching is investigated by analyzing movements of eye with and without smell using a detector. Visual image of foods is presented to a viewer, and his or her gazing time on a food object is detected. This experiment reveals advertisement accompanied with smell is more attracted to a viewer. The proposed method for smell presentation is to induce a person inhalation by tactile sensation caused by airflow and released smell at a time of inhalation. This experiment discusses the possibility of inducing subjects to receive smell efficiently.
Download
Paper Nr: 42
Title:

Architectural Model for Visualization of High Definition Images on Mobile Devices
Authors:

Germán Corredor, Daniel Martínez, Eduardo Romero and Marcela Iregui
Abstract: In recent years, the mobile device demand has largely increased because of the accessibility, ubiquity and portability of such devices, which are being used not only for personal purposes but also in several applications like education, science, entertainment, commerce and industry, among others. Visualization and interaction with high definition multimedia content, like large images and videos, using mobile devices, represents a challenge because of their very limited machine resources and bandwidth. For such application, this content requires special treatment so that users can properly access and interact. In this article, it is proposed an architectural model for efficient streaming and visualization of very large images on mobile devices using the JPEG2000 standard and an adapted image transfer protocol. Results show that the introduced architecture is effective for visualizing regions of large images and presents good performance, both for transmission and decoding processes, allowing a simple and dynamic interaction between user and images.
Download
Paper Nr: 47
Title:

Consumer Propensity and Location Analysis based Real-time Location Tracing Advertisement Service Design and Implementation - Real-time Location based Advertisement System
Authors:

Daehee Won, Yoonsoo Kim, Hangki Joh, Intae Ryoo and Doungyung Suh
Abstract: While distributing Android free of charge, Google intended to expose its advertisements on the platforms to seize users’ eyes and make profits. However, smart phones are kept in bags or pockets during most of the time instead of showing screens in front of users’ eyes. If the time during which users’ eyes cannot be seized becomes longer advertisement effects will decrease as much. In this study, in order to solve these problems, consumers’ movement paths are grasped using continuous screens based on the results of analyses of consumer propensity to replay advertisement images. Advertisement image replay lists are composed of related advertisements based on the key words set by consumers. The relevant project was named as shADow meaning Advertisements that follow like shadows.
Download
Paper Nr: 56
Title:

Autonomous Constructing Everyday Projected Displays
Authors:

Cui Xie, Qi Wang and Wei Cheng
Abstract: This paper presents an autonomous geometric correction method to support constructing a projector-based large display for everyday use, which includes offline and online processing phase. The offline process is focus on an automatic, fast and robust approach for the geometric registration of projector-camera system. The online stage is mainly the implementing real-time image warping via modern graphics hardware to achieve the final corrected images without first acquiring geometric information of the screen’s surface. Since a simple checkerboard pattern is used to facilitate building the mapping of the corner correspondences of projector-camera image, and a perspective projection invariant rational Bezier patch is used to approximately represent the mapping, our method saves a lot of computing time and become easier and robust. Therefore, the achieved transformation can be used for online image warping for actual projection. As a result, a layman user can get a corrected image displayed on a non-planar surface for the point of view of the camera easily.
Download
Paper Nr: 77
Title:

Content Meets Semantics: Smarter Exploration of Image Collections - Presentation of Relevant Use Cases
Authors:

Ilaria Bartolini
Abstract: Current techniques for the management of image collections exploit either user-provided annotations or automatically-extracted visual features. Although effective, the approach based on annotations cannot be efficient since the manual process of data tagging prevents its scalability. On the other hand, the organization and search grounded on visual features, such as color and texture, is known to be a powerful (since it can be made fully automatic), yet imprecise, retrieval paradigm, because of the semantic gap problem. This position paper advocates the combination of visual content and semantics as a critical binomial for effectively and efficiently managing and browsing image databases satisfying users’ expectations in quickly locating images of interest.
Download
Paper Nr: 79
Title:

Next Generation TV through Automatic Multimedia Annotation Systems - A Hybrid Approach
Authors:

Joël Dumoulin, Marco Bertini, Alberto Del Bimbo, Elena Mugellini, Omar Abou Khaled and Maria Sokhn
Abstract: After the advent of smartphones, it is time for television to see its next big evolution, to become smart TVs. But to provide a richer television user experience, multimedia content first has to be enriched. In recent years, the evolution of technology has facilitated the way to take and store multimedia assets, like photographs or videos. This causes an increased difficulty in multimedia resources retrieval, mainly because of the lack of methods that handle non-textual features, both in annotation systems and search engines. Moreover, multimedia sharing websites like Flickr or YouTube, in addition to information provided by Wikipedia, offer a tremendous source of knowledge interesting to be explored. In this position paper, we address the automatic multimedia annotation issue, by proposing a hybrid system approach. We want to use unsupervised methods to find relationships between multimedia elements, referred as hidden topics, and then take advantage of social knowledge to label these resulting relationships. Resulting enriched multimedia content will allow to bring new user experience possibilities to the next generation television, allowing for instance the creation of recommender systems that merge this information with user profiles and behavior analysis.
Download
Paper Nr: 80
Title:

Centroid-based Clustering for Student Models in Computer-based Multiple Language Tutoring
Authors:

Maria Virvou, Efthymios Alepis and Christos Troussas
Abstract: This paper proposes an approach for the initialization and the construction of student models in an intelligent tutoring system that teaches multiple foreign languages. The basic concept for the construction of the initial user models is to assign each new student to a model with similar characteristics. As it is quite easy to understand that a tutoring system has rather little information about its new users, our effort is to provide as much information as possible for each specific user relying on the user’s initial data. To this end, a machine learning algorithm, namely k-means, is responsible for creating clusters relying on the system’s pre-entered past data and as a next step, each new entry is assigned to the nearest centroid.
Download
Paper Nr: 81
Title:

Influence of Different Phoneme Mappings on the Recognition Accuracy of Electrolaryngeal Speech
Authors:

Petr Stanislav and Josef V. Psutka
Abstract: This paper presents the initial steps towards building speech recognition system that is able to efficiently process electrolaryngeal substitute speech produced by laryngectomees. Speakers after total laryngectomy are characterized by restricted aero-acoustic properties in comparison with normal speakers and their speech is therefore far less intelligible. We suggested and tested several approaches to acoustic modeling within the ASR system that would be able to cope with this lower intelligibility. Comparative experiments were also performed on the healthy speakers. We tried several mappings that unify unvoiced phonemes with their voiced counterparts in the acoustic modeling process both on monophone and triphone level. Systems using zerogram and trigram language models were evaluated and compared in order to increase the credibility of the results.
Download
Paper Nr: 82
Title:

Large Scale Similar Song Retrieval using Beat-aligned Chroma Patch Codebook with Location Verification
Authors:

Yijuan Lu and Joseph E. Cabrera
Abstract: With the popularity of song search applications on Internet and mobile phone, large scale similar song search has been attracting more and more With the popularity of song search applications on Internet and mobile phone, large scale similar song search has been attracting more and more attention in recent years. Similar songs are created by altering the volume levels, timing, amplification, or layering other songs on top of an original song. Given the large scale of songs uploaded on the Internet, it is demanding but challenging to identify these similar songs in a timely manner. Recently, some state-of-the-art large scale music retrieval approaches represent songs with a bag of audio words by quantizing local features, such as beat-chroma patches, solely in the feature space. However, feature quantization reduces the discriminative power of local features, which causes many false audio words matches. In addition, the location clues among audio words in a song is usually ignored or exploited for full location verification, which is computationally expensive. In this paper, we focus on similar song retrieval, and propose to utilize beat-aligned chroma patches for large scale similar song retrieval and apply location coding scheme to encode the location relationships among beat-aligned chroma patches in a song. Our approach is both efficient and effective to discover true matches of beat chroma patches between songs with low computational cost. Experiments in similar songs search on a large song database reveal the promising results of our approach.
Download
Paper Nr: 85
Title:

A New Tool for the Analysis of Heart Rate Variability of Long Duration Records
Authors:

Ricardo Chorão, Joana Sousa, Tiago Araújo and Hugo Gamboa
Abstract: The increased masses of data confronting us, originate a pressing need for the creation of a user interface for better handling and extracting knowledge from it. In this work we developed such a tool for the analysis of Heart Rate Variability (HRV). The analysis of HRV in patients with neuromuscular diseases, sleep disorders and cardiorespiratory problems has a strong impact on clinical practice. It has been widely used for monitoring the autonomic nervous system (ANS), whose regulatory effect controls the cardiac activity. These patients need to be continuously monitored, which originates data with huge sizes. Our interactive tool can perform a fast analysis of HRV from such data. It provides the analysis of HRV in time and frequency domains, and from non-linear methods. The tool is suitable to be run in a web environment, rendering it highly portable. It includes a programming feature, which enables the user to perform additional analysis of the data by giving direct access to the signals in a signal processing programming environment. We also added a report generation functionality, which is extremely important from a clinical standpoint, on which the evolution in time of relevant HRV parameters is depicted.
Download
Paper Nr: 7
Title:

Computer Games Sound Effects - Recording, Postproduction and Existing Database
Authors:

Bartosz Ziółko, Martyna Gromotka and Mariusz Ziółko
Abstract: The paper describes the process of building a new database of sound effects recordings for computer games and the first version of such product. Ways of applying signal processors for postproduction is described, as well as differences in audio edition for films and games. Some aspects of using sounds in games are also mentioned as well as the first version of the list of possible tags of the audio files in the database. Both the language of the tags and the datatabse will be substanially enlarged.
Download
Paper Nr: 17
Title:

Adaptive Visualization of Segmented Digital Ink Texts in Chinese based on Context
Authors:

Xi-Wen Zhang, Hao Bai and Yong-Gang Fu
Abstract: Digital ink texts in Chinese can neither be converted into users’ desired layouts nor be recognized until they are segmented correctly. There are many errors in automatically segmented results because the texts are free forms and mixed with other languages, as well as their Chinese characters have small gaps and complex structures. Paragraphs, text lines, and characters (recognizable language symbols) may be wrongly extracted. It is a prerequisite to visualize segmented results for further correcting wrong extracted objects using human-computer interaction. Thus, an adaptive approach based on context is proposed to visualize segmented digital ink texts in Chinese. Each extracted object is adaptively visualized by shape and colour labels according to relations between it and its neighbours. Confidences of extracted objects are also visualized with bounding shapes with different line widths. Each object’s contexts are constructed from it and other objects invoked by it, where an optimum visualization is identified. We have conducted experiments using real-life segmented digital ink texts in Chinese and compared the proposed approach with others. Experimental results demonstrate that the proposed approach is feasible, flexible, and effective.
Download
Paper Nr: 48
Title:

Semi-dynamic Calibration for Eye Gaze Pointing System based on Image Processing
Authors:

Kohichi Ogata and Kohei Matsumoto
Abstract: In this paper, we propose two semi-dynamic calibration methods for compensating for user’s head movements for an eye gaze pointing system. Since the user perceives degradation in pointing accuracy during use, an effective compensatory calibration by the user which does not require additional apparatus or high cost calculation can be a useful solution for the problem. The proposed semi-dynamic calibration methods lead the user to gaze at 1 or 3 points on the computer screen and reduce the gap between the true eye gaze direction and the position of the mouse pointer. Experimental results showed that the proposed methods were capable of pointing the mouse pointer within 20 pixels at a distance of about 60 cm between the user and the display.
Download
Paper Nr: 76
Title:

Simulated Annealing based Parameter Optimization of Time-frequency e-filter Utilizing Correlation Coefficient
Authors:

Tomomi Matsumoto, Mitsuharu Matsumoto and Shuji Hashimoto
Abstract: Time-Frequency e-filter (TF e-filter) can reduce different types of noise from a single-channel noisy signal while preserving the signal that varies drastically such as a speech signal. It can reduce not only small stationary noise but also large nonstationary noise. However, it has some parameters whose values are set empirically. So far, there are few studies to optimize the parameter of TF e-filter automatically. In this paper, we employ the correlation coefficient of the filter output and the difference between the filter input and output as the evaluation function of the parameter optimization. We also propose an algorithm to set the optimal parameter of TF e-filter automatically. The experimental results show that we can obtain the adequate parameter in TF e-filter automatically by using the proposed method.
Download