Chang-Bae Moon

Audio Recorder Identification Using Reduced Noise Features

In this paper, we propose an audio recorder identification method as one of digital forensic technologies, where Wiener filter is used to extract noise sounds of recorders and their features are extracted by MIRtoolbox. A recorder identification model is generated by training SVM with the extracted noise features. To improve the identification performance, a new feature reduction method which uses inter-classes standard deviations of features is adopted. The experimental results for 11 audio recorders show 1% improvement over the method with no feature reduction. The improvement is not too noticeable as expected, but the number of features can be reduced up to one third of the method with no feature reduction. The results also show that the proposed feature reduction method is competitive over the other well-known methods such as PCA, LDA and R-squared.

Augmented Reality Logo System Based on Android platform

A mobile phone is becoming no longer a voice communication tool due to smartphones and mobile internet. Also, it now becomes a total entertainment device on which we can play game and get services by variety applications through the Web. As smartphones are getting more popular, their usages are also increased, which makes the interest of advertising industry in mobile advertisement increased but it is bound to be limited by the size of the screen. In this paper, we suggest an augmented reality logo system based on Android platform to maximize the effect of logo advertisement. After developing software and mounting it on a real smartphone, its performances are analyzed in various ways. The results show the possibility of its application to real world but it`s not enough to provide real time service because of the low performance of hardware.

Performance Evaluation of a Generalized Music Mood Classification Model

This paper focuses on building a generalized mood classification model with many mood classes instead of a personalized one with few mood classes. Two methods are adopted to improve the performance of mood classification. The one of them is a new feature reduction based on standard deviation of feature values, which is designed to solve the problem of lowered performance when all 391 features provided by MIR toolbox used to extract features of music. Our experiments show that the feature reduction method suggested in this paper has better performance than that of the conventional dimension reduction methods, R-Squared and PCA. As performance improvement by feature reduction only is limited, the modular neural network approach is used additionally. The experiments also show that the method improves the performance effectively.

LCD BLU Defects Detection System with Sidelight

A Cold Cathode Fluorescent Lamp(CCFL) is used as a LCD Monitor`s backlight widely. The most common way to check CCFL`s defects is an examination with the naked eye. This naked eye examination can cause examination inconsistencies and industrial disasters. A shooting environment and detection algorithms are important for finding CCFL defects automatically. This paper presents CCFL defect detection algorithms using images captured under the shooting environment with sidelight which is one of the shooting environment we have suggested. The experimental result shows 4.65% of overdetection and 5.37% of unsuccessful defect detection of CCFL.