• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.583-586
• /
• 1999

In this paper, the wavelet transform is performed in the input 256$\times$256 color image and decomposes a image into low-pass and high-pass components. Since the high-pass band contains the components of three directions, edges are detected by combining three parts. After finding the position of face using the histogram of the edge component, a face region in low-pass band is cut off. Since RGB color image is sensitively affected by luminances, the image of low pass component is normalized, and a facial region is detected using face color informations. As the wavelet transform decomposes the detected face region into three layer, the dimension of input image is reduced. In this paper, we use the 3000 images of 10 persons, and KL transform is applied in order to classify face vectors effectively. FCM(Fuzzy C-Means) algorithm classifies face vectors with similar features into the same cluster. In this case, the number of cluster is equal to that of person, and the mean vector of each cluster is used as a codebook. We verify the system performance of the proposed algorithm by the experiments. The recognition rates of learning images and testing image is computed using correlation coefficient and Euclidean distance.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.39 no.1
• /
• pp.32-39
• /
• 2002

Vector quantization for image compression needs expensive encoding time to find the closest codeword to the input vector. This paper proposes a search algorithm for fast vector quantization encoding. Firstly, we derive a robust condition based on the efficient topological structure of the codebook to dramatically eliminate unnecessary matching operations from the search procedure. Then, we Propose a fast search algorithm using the elimination condition. Simulation results show that with little preprocessing and memory cost, the encoding time of the proposed algorithm is reduced significantly while the encoding quality remains the same with respect to the full search algorithm. It is also found that the Proposed algorithm outperforms the existing search algorithms.

• Annual Conference on Human and Language Technology
• /
• 2001.10d
• /
• pp.209-212
• /
• 2001

본 논문에서는 text-to-speech 시스템에서 사용할 억양 모델을 위해 벡터 양자화(vector quantization) 방식을 이용한다. 어절 경계강도(break index)는 세단계로 분류하였고, CART(Classification And Regression Tree)를 사용하여 어절 경계강도의 예측 규칙을 생성하였다. 예측된 어절 경계강도를 바탕으로 운율구를 예측하였으며 운율구는 다섯 개의 억양 패턴으로 분류하였다. 하나의 운율구는 정점(peak)의 시간축, 주파수축 값과 이를 기준으로 한 앞, 뒤 기울기를 추출하여 네 개의 파라미터로 단순화하였다. 운율구에 대해서 먼저 운율구가 문장의 끝일 경우와 아닐 경우로 분류하고, 억양 패턴 다섯 개로 분류하여. 모두 10개의 운율구 set으로 나누었다. 그리고 네 개의 파라미터를 가지고 있는 운율구의 억양 패턴을 벡터 양자화 방식을 이용하여 분류(clusteing)하였다 운율의 변화가 두드러지는 조사와 어미는 12 point의 기본주파수 값을 추출하고 벡터 양자화하였다. 운율구와 조사 어미의 codebook index는 문장에 대한 특징 변수 값을 추출하고 CART를 사용하여 예측하였다. 합성할 때에는 입력 tort에 대해서 운율구의 억양 파라미터를 추정한 다음, 조사와 어미의 12 point 기본주파수 값을 추정하여 전체 억양 곡선을 생성하였고 본 연구실에서 제작한 음성합성기를 통해 합성하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.12
• /
• pp.5170-5188
• /
• 2015

This paper proposes a novel progressive secret image-hiding scheme based on the inpainting technique, the vector quantization technique (VQ) and the exploiting modification direction (EMD) technique. The proposed scheme first divides the secret image into non-overlapping blocks and categorizes the blocks into two groups: complex and smooth. The blocks in the complex group are compressed by VQ with PCA sorted codebook to obtain the VQ index table. Instead of embedding the original secret image, the proposed method progressively embeds the VQ index table into the cover images by using the EMD technique. After the receiver recovers the complex parts of the secret image by decoding the VQ index table from the shadow images, the smooth parts can be reconstructed by using the inpainting technique based on the content of the complex parts. The experimental results demonstrate that the proposed scheme not only has the advantage of progressive data hiding, which involves more shadow images joining to recover the secret image so as to produce a higher quality steganography image, but also can achieve high hiding capacity with acceptable recovered image quality.

• The Journal of the Acoustical Society of Korea
• /
• v.19 no.1
• /
• pp.67-72
• /
• 2000

Particularly based on the signal bias removal(SBR) algorithm for compensating the corrupted speech, this paper presents a new algorithm which is independent of environments, minimizes the amount of computation, and is readily applicable to the conventional recognition system. To this end, a multiple-bias algorithm and a partial codebook search algorithm have been added to the conventional SBR algorithm. The simulation results show that combining the two algorithms proposed in this paper provides a reduction of computation time to 1/8 times as well as an improvement of the recognition rate from 77.58% of the conventional system to 81.32%.

• The Journal of the Acoustical Society of Korea
• /
• v.13 no.1E
• /
• pp.89-95
• /
• 1994

This paper describes development of a real-time voice dialing system which can recognize around one hundred word vocabularies in speaker independent mode. The voice recognition algorithm in this system is implemented on a DSP board with a telephone interface plugged in an IBM PC AT/486. In the DSP board, procedures for feature extraction, vector quantization(VQ), and end-point detection are performed simultaneously in every 10 msec frame interval to satisfy real-time constraints after detecting the word starting point. In addition, we optimize the VQ codebook size and the end-point detection procedure to reduce recognition time and memory requirement. The demonstration system has been displayed in MOBILAB of the Korean Mobile Telecom at the Taejon EXPO'93.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.21-24
• /
• 2001

본 논문은 신경 회로망을 이용하여, 전화망 대역의 음성, 즉, narrowband 음성에서 wideband 음성을 복원하고자 했다. BP 알고리즘을 사용하는 기존의 신경 회로망의 경우에는 음성과 같이 복잡하고 크기가 큰 훈련데이터에 대해서는 훈련이 제대로 되지 않는 단점이 있다. 그러므로 븐 논문에서는 이를 해결하기 위해 입력으로 들어온 LPC 켑스트럼 벡터를 k-means 알고리즘을 이용하여 미리 정한 개수의 cluster로 나눈 다음, 각각의 cluster에 대해 독립적인 신경 회로망을 적용했다 이로 인해 각각의 신경 회로망은 제한되고 서로 상관관계가 많은 음성들만 훈련하면 되므로, 기존의 신경 회로망에서 생기는 훈련의 정체를 개선할 수 있었다. 또 clustering 과정에서 생기는 오류를 보완하기 위해 후보신경 로망들의 출력에 fuzzy 개념을 적용해서 최종 출력을 내도록 했다 실험 결과에서, 제안한 알고리즘은 기존의 codebook mapping 알고리즘보다 스펙트럼 거리척도에 의한 비교 및 주관적인 음질 평가 양쪽에서 개선된 성능을 보였다.

• Journal of Power Electronics
• /
• v.16 no.2
• /
• pp.760-768
• /
• 2016

Three-phase grid-connected inverters are widely applied in the fields of new energy power generation, electric vehicles and so on. Islanding detection is necessary to ensure the stability and safety of such systems. In this paper, feature recognition technology is applied and a novel islanding detection method is proposed. It can identify the features of inverter systems. The theoretical values of these features are defined as codebooks. The difference between the actual value of a feature and the codebook is defined as the quantizing distortion. When islanding happens, the sum of the quantizing distortions exceeds the threshold value. Thus, islanding can be detected. The non-detection zone can be avoided by choosing reasonable features. To accelerate the speed of detection and to avoid miscalculation, an active islanding detection method based on feature recognition technology is given. Compared to the active frequency or phase drift methods, the proposed active method can reduce the distortion of grid-current when the inverter works normally. The principles of the islanding detection method based on the feature recognition technology and the improved active method are both analyzed in detail. An 18 kVA DSP-based three-phase inverter with the SVPWM control strategy has been established and tested. Simulation and experimental results verify the theoretical analysis.

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10c
• /
• pp.384-386
• /
• 1998

The use of hand gesture provides an attractive alternative to cumbersome interface devices for human-computer interaction(HCI). Many methods hand gesture recognition using visual analysis have been proposed such as syntactical analysis, neural network(NN), Hidden Markov Model(HMM) and so on. In our research, a HMMs is proposed for alphabetical hand gesture recognition. In the preprocessing stage, the proposed approach consists of three different procedures for hand localization, hand tracking and gesture spotting. The hand location procedure detects the candidated regions on the basis of skin-color and motion in an image by using a color histogram matching and time-varying edge difference techniques. The hand tracking algorithm finds the centroid of a moving hand region, connect those centroids, and thus, produces a trajectory. The spotting a feature database, the proposed approach use the mesh feature code for codebook of HMM. In our experiments, 1300 alphabetical and 1300 untrained gestures are used for training and testing, respectively. Those experimental results demonstrate that the proposed approach yields a higher and satisfying recognition rate for the images with different sizes, shapes and skew angles.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.40 no.8
• /
• pp.1469-1476
• /
• 2015

We propose a beam selection algorithm that improves inter sector SIR using a code-book of a circular array antenna in multi-sector wireless mesh network environments. The proposed method improves SIR using a combination of fed back code-book and guarantees QoS of all nodes. Computer simulation exhibits the proposed scheme demonstrates 4.42dB higher SIR than that of the conventional code-book method, QoS with proportional fair is improved by 1.70dB and fact that all nodes are satisfied Qos is also shown.