• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.40 no.4
• /
• pp.286-291
• /
• 2003

In this Paper, we designed an LSP(Line Spectral Pairs) parameter quantizer with cascaded structure of vector quantizer and scalar quantizer for the wideband speech coder. We have chosen the 16th-order of the LP coefficients. These coefficients are then transformed into the LSP parameters which have the excellent properties for quantization and easy stability checking condition of synthesis filter. In the first stage of quantization, input LSP parameters are split-vector-quantized using two 8-th order codebooks. In the second stage, the components of residual vector are individually quantized by the scalar quantizer utilizing the ordering property of LSP parameters. The designed adaptive VQ-SQ quantizer using 35 bits/frame shows the wideband transparency that the average spectral distortion should be less than 1.6 ㏈ and less than 4% of the frames should have SD above 3 ㏈. The simulation results show that the designed quantizer provides a 2-3 bits/frame saving over the typical vector-scalar quantizer.

• Journal of the Institute of Convergence Signal Processing
• /
• v.23 no.4
• /
• pp.194-199
• /
• 2022

Preprocessing for high-quality data is required for high accuracy and usability in various and complex image data-based industries. However, when a contaminated hostile example that combines noise with existing image or video data is introduced, which can pose a great risk to the company, it is necessary to restore the previous damage to ensure the company's reliability, security, and complete results. As a countermeasure for this, restoration was previously performed using Defense-GAN, but there were disadvantages such as long learning time and low quality of the restoration. In order to improve this, this paper proposes a method using adversarial examples created through FGSM according to image segmentation in addition to using the VQ-VAE model. First, the generated examples are classified as a general classifier. Next, the unsegmented data is put into the pre-trained VQ-VAE model, restored, and then classified with a classifier. Finally, the data divided into quadrants is put into the 4-split-VQ-VAE model, the reconstructed fragments are combined, and then put into the classifier. Finally, after comparing the restored results and accuracy, the performance is analyzed according to the order of combining the two models according to whether or not they are split.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.47 no.4
• /
• pp.69-78
• /
• 2010

The irreversible 9/7 JPEG2000, which is one of sub-band coding techniques, has a problem of severe picture quality distortion at the edge and the background caused by the quantization error below 0.15bpp. In this paper, to solve such problems we propose a VQ(Visual Quantizer) based on L-pdf(Laplace probability density function) statistical characteristics of high frequency sub-bands. The proposed VQ is designed by visual parameter for improving the subjective quality and weighting parameter for increasing the compression ratio. A proposed method, based on 9/7 JPEG2000 scheme, gives the high subjective quality to reconstructed images below 0.15bpp and provides minimum MSE(Mean-Squared Error) regardless of the compression ratio.

• The Journal of the Acoustical Society of Korea
• /
• v.14 no.1E
• /
• pp.46-55
• /
• 1995

In this paper, we present a new wavelet transform-based CELP coder, called band selection wavelet transform CELP (BS-WTCELP) operated at 4.8 kbps. The proposed algorithm uses a band selection scheme of frequency bands of wavelet transform and selective vector quantization (VQ). The band selection and selective VQ structure is implemented by using a classified VQ structure. The proposed algorithm has about 0.5-1.0 dB improvement in segmental SNR compared with the conventional CELP that uses the random codebook search, while is has significantly reduced computational and storage complexity. Many experimental results have shown that the proposed algorithm is more suitable for most real-applications than the conventional CELP and wavelet transform CELP.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.9C
• /
• pp.877-886
• /
• 2005

Various frame recovery algorithms have been suggested to overcome the communication quality degradation problem due to Internet-typical impairments on Voice over IP(VoIP) communications. In this paper, we propose a new receiver-based recovery method which is able to enhance recovered speech quality with almost free computational cost and without an additional increment of delay and bandwidth consumption. Most conventional recovery algorithms try to recover the lost or erroneous speech frames by reconstructing missing coefficients or speech signal during speech decoding process. Thus they eventually need to modify the decoder software. The proposed frame recovery algorithm tries to reconstruct the missing frame itself, and does not require the computational burden of modifying the decoder. In the proposed scheme, the Vector Quantization(VQ) codebook indices of the erased frame are directly estimated by referring the pre-computed VQ Codebook Index Interpolation Tables(VCIIT) using the VQ indices from the adjacent(previous and next) frames. We applied the proposed scheme to the ITU-T G.723.1 speech coder and found that it improved reconstructed speech quality and outperforms conventional G.723.1 loss recovery algorithm. Moreover, the suggested simple scheme can be easily applicable to practical VoIP systems because it requires a very small amount of additional computational cost and memory space.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.4
• /
• pp.685-690
• /
• 1994

A problem associated with vector quantization(VQ) is the computational complexity incurred in searching for a codevector with the closet to a given input vector, where the complexity increases exponentionally with proportion to codebook size and then limits practical application. In this paper, a simple and fast, but efficient, VQ encoding algorithm is presented using a reference codevector as start codevector of premature exit condition, which eliminates distance claculation of unlikely codevectors. The algorithm is to find reference codevector having the possibility to be the nearest vector to input vector first and then to incorporate premature exit condition. The proposed algorithm needs only 10~15% of mathematical operations compared with the conventional full search VQ. Algorithm the number of additions and comparsions of the proposed algorithm is not reduced greatly, the number of multiplication is reduced up to 70~80% compared with other fast VQ encoding methods.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2000.11a
• /
• pp.467-470
• /
• 2000

Vector quantization (VQ) is widely used in many high-quality and high-rate data compression applications such as speech coding, audio coding, image coding and video coding. When the size of a VQ codebook is large, the computational complexity for the full codeword search method is a significant problem for many applications. A number of complexity reduction algorithms have been proposed and investigated using such properties of the codebook as the triangle inequality. This paper proposes a new structured VQ search algorithm that is based on a multi-stage structure for searching for the best codeword. Even using only two stages, a significant complexity reduction can be obtained without any loss of quality.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1998.06c
• /
• pp.95-98
• /
• 1998

본 논문에서는 이산적인 HMM(Hidden Markov Model)을 이용한 고립단어 인식 시스템에서 입력특징 벡터의 변별력을 향상시키기 위해 수정된 집단화 알고리듬을 제안하므로써 K-means나 LBG 알고리듬을 이용한 기존의 HMM에 비해 2.16%의 인식율을 향상시켰다. 또한 HMM학습과정에서 불충분한 학습데이타로 인해 발생되는 인식율저하의 문제를 해소하기 위해 개선된 smoothing 기법을 제안하므로써 화자독립 실험에서 3.07%의 인식율을 향상시켰다. 본 논문에서 제안한 두가지 알고리듬을 모두 적용하여 최종적으로 실험한 VQ/HMM에서는 기존의 방식에 비해 화자독립 인식실험 결과 평균 인식율이 4.66% 개선되었다.

• The Journal of the Acoustical Society of Korea
• /
• v.17 no.4
• /
• pp.72-75
• /
• 1998

본 논문에서는 기존에 제시된 수정된 K-평균 방법을 이용한 VQ 학습 알고리즘을 분석하고, 보다 개선된 성능을 보이는 학습 알고리즘을 제안한다. 수정된 K-평균 학습 알고 리즘은 자기 집단에 속하는 데이터의 중심을 데이터의 중심을 새로운 코드워드로 삼는 것이 아니라 현재 코드워드와 새로 구한 집단의 중심을 연결한 선상에서 새로 구한 중심 너머의 일정한 점을 새로운 코드워드로 선택하는 방식이다. 본 논문에서는 이렇게 구한 새로운 코 드워드가 어떠한 조건을 만족할 때 알고리즘이 반복적 감소의 성질을 가지는지 살펴보고, 그 조건을 만족시키는 영역 중 기존의 방식보다 더 좋은 성능을 보이는 코드워드 선택법을 제시함으로써 개선된 학습 알고리즘을 제안한다.

• Speech Sciences
• /
• v.8 no.3
• /
• pp.143-148
• /
• 2001

We propose a new scheme for VQ codebook search. The procedure is in between the binary-tree-search and full-search and thus might be called N-ary search of a codebook. Through the experiment performed on 7200 frames spoken by 25 speakers, we confirmed that the best codewords as good as by the full-search were obtained at moderate time consumption comparable to the binary-tree-search. In application to speech recognition by HMM/VQ with Bakis model, where appearance of a specific codeword is essential in the parameter training phase, the method proposed here is expected to provide an efficient training procedure.