• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.8 no.9
• /
• pp.265-273
• /
• 2018

In this paper, we propose an acoustic event detection method in tunnels using non-negative tensor factorization (NTF) and hidden Markov model (HMM) applied to multi-channel audio signals. Incidents in tunnel are inherent to the system and occur unavoidably with known probability. Incidents can easily happen minor accidents and extend right through to major disaster. Most incident detection systems deploy visual incident detection (VID) systems that often cause false alarms due to various constraints such as night obstacles and a limit of viewing angle. To this end, the proposed method first tries to separate and detect every acoustic event, which is assumed to be an in-tunnel incident, from noisy acoustic signals by using an NTF technique. Then, maximum likelihood estimation using Gaussian mixture model (GMM)-HMMs is carried out to verify whether or not each detected event is an actual incident. Performance evaluation shows that the proposed method operates in real time and achieves high detection accuracy under simulated tunnel conditions.

• Proceedings of the KSR Conference
• /
• 2011.05a
• /
• pp.1161-1166
• /
• 2011

A camera is mounted vertically downwards viewing the people heads from the top. This configuration is successful in people counting technique especially when only a few isolated people pass through a counting region in a non-crowded situation. Thus, this paper describes object counting which detects and count moving people using mixture of gaussian and motion vector. This method is intended to estimates the number of people in outdoor environment. This method use single gaussian background modeling which is more robust an noise and has adaptiveness. The experimental results that is based on mixture of gaussian and motion vector is also helpful to design intelligent surveillance.

• Korean Journal of Remote Sensing
• /
• v.19 no.6
• /
• pp.465-478
• /
• 2003

This paper presents two methods for automatic estimation of threshold values in unsupervised change detection of multi-temporal remote sensing images. The proposed methods consist of two analytical steps. The first step is to compute the parameters of a 3-component Gaussian mixture model from difference or ratio images. The second step is to determine a threshold value using Bayesian rule for minimum error. The first method which is an extended version of Bruzzone and Prieto' method (2000) is to apply an Expectation-Maximization algorithm for estimation of the parameters of the Gaussian mixture model. The second method is based on an iterative thresholding algorithm that successively employs thresholding and estimation of the model parameters. The effectiveness and applicability of the methods proposed here were illustrated by two experiments and one case study including the synthetic data sets and KOMPSAT-1 EOC images. The experiments demonstrate that the proposed methods can effectively estimate the model parameters and the threshold value determined shows the minimum overall error.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.7
• /
• pp.528-538
• /
• 2003

A SCCRS (Speech and Character Combined Recognition System) is developed for working on mobile devices such as PDA (Personal Digital Assistants). In SCCRS, the feature extraction is separately carried out for speech and for hand-written character, but the recognition is performed in a common engine. The recognition engine employs essentially CHMM (Continuous Hidden Markov Model), which consists of variable parameter topology in order to minimize the number of model parameters and to reduce recognition time. For generating contort independent variable parameter model, we propose the SSMS(Successive State and Mixture Splitting), which gives appropriate numbers of mixture and of states through splitting in mixture domain and in time domain. The recognition results show that the proposed SSMS method can reduce the total number of GOPDD (Gaussian Output Probability Density Distribution) up to 40.0% compared to the conventional method with fixed parameter model, at the same recognition performance in speech recognition system.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.6
• /
• pp.49-56
• /
• 1999

In many continuous speech recognition systems based on HMMs, decision tree-based state tying has been used for not only improving the robustness and accuracy of context dependent acoustic modeling but also synthesizing unseen models. To construct the phonetic decision tree, standard method performs one-level pruning using just single Gaussian triphone models. In this paper, two novel approaches, two-level decision tree and multi-mixture decision tree, are proposed to get better performance through more accurate acoustic modeling. Two-level decision tree performs two level pruning for the state tying and the mixture weight tying. Using the second level, the tied states can have different mixture weights based on the similarities in their phonetic contexts. In the second approach, phonetic decision tree continues to be updated with training sequence, mixture splitting and re-estimation. Multi-mixture Gaussian as well as single Gaussian models are used to construct the multi-mixture decision tree. Continuous speech recognition experiment using these approaches on BN-96 and WSJ5k data showed a reduction in word error rate comparing to the standard decision tree based system given similar number of tied states.

• The Journal of the Korea Contents Association
• /
• v.20 no.3
• /
• pp.424-431
• /
• 2020

This paper focused on the development of sports robot that detects a human player and shots a serve ball automatically. When robot technologies apply to the sports machine, the domain problems occurs such as outdoor environments and playing condition to recognize the visual and the vocal modalities. Gaussian mixture model and Kalman filter are used to detect the player's position in the left, right, and depth direction and to avoid the noises caused by the player's posture variation around the net. The sports robot is designed by the pan-tilt structure to shot a serve ball by pneumatic control under the multi layered software architecture. Finally, the proposed tracking and the machine performance are discussed by experimental results.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.9 no.9
• /
• pp.977-982
• /
• 2014

In this paper, we propose a system which can detect the flame in real time from the high-quality IP camera. First, open directly the RTSP streams transmitted from the IP camera using the library FFmpeg as opening a video file. The second thing is to extract the background images from video signal using Gaussian mixture model. Then the foreground images are obtained through subtracting operation between the input image and the background image. Separated foreground image through a mathematical morphology operation are considered as candidate area. By analysing colour information and dynamic characteristics of the candidate area, flame is determined finally. Through the experiments with input videos from IP camera, the proposed algorithms were useful to detect flames.

• Journal of Digital Contents Society
• /
• v.14 no.1
• /
• pp.117-124
• /
• 2013

This paper suggests the robust method of extraction for moving objects in illumination variation by using image sequence from an immovable camera. The most difficult part of the implication is the effect by illumination and noise. The object area is hardly estimated when the dusky area occurs in illumination variation by time change. This thesis describes the extraction of moving objects employed by Gaussian mixture model which is noise robust measure. Also, the report suggests the elimination method of illumination part in input image by the representative illumination image which is defined to minimize the illumination influence.

• The Journal of the Acoustical Society of Korea
• /
• v.24 no.1
• /
• pp.58-68
• /
• 2005

In this paper, we propose a scheme to improve discriminative property in the feature compensation method for robust speech recognition under noisy environments. The estimation of noisy speech model used in existing feature compensation methods do not guarantee the computation of posterior probabilities which discriminate reliably among the Gaussian components. Estimation of Posterior probabilities is a crucial step in determining the discriminative factor of the Gaussian models, which in turn determines the intelligibility of the restored speech signals. The proposed scheme employs minimum classification error (MCE) training for estimating the parameters of the noisy speech model. For applying the MCE training, we propose to identify and determine the 'competing components' that are expected to affect the discriminative ability. The proposed method is applied to feature compensation based on parallel combined mixture model (PCMM). The performance is examined over Aurora 2.0 database and over the speech recorded inside a car during real driving conditions. The experimental results show improved recognition performance in both simulated environments and real-life conditions. The result verifies the effectiveness of the proposed scheme for increasing the performance of robust speech recognition systems.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.7 no.6
• /
• pp.1293-1299
• /
• 2012

In this paper, a real-time vehicle detection and tracking algorithms is proposed. The vehicle detection could be processed using GMM (Gaussian Mixture Model) algorithm and mathematical morphological processing with HD CCTV camera images. The vehicle tracking based on separated vehicle object was performed using Euclidean distance between detected object. In more detail, background could be estimated using GMM from CCTV input image signal and then object could be separated from difference image of the input image and background image. At the next stage, candidated objects were reformed by using mathematical morphological processing. Finally, vehicle object could be detected using vehicle size informations dependent on distance and vehicle type in tunnel. The vehicle tracking performed using Euclidean distance between the objects in the video frames. Through computer simulation using recoded real video signal in tunnel, it is shown that the proposed system works well.