• 한국통신학회논문지
• /
• 제23권9A호
• /
• pp.2355-2364
• /
• 1998

In conventional fractal decoding procedure, the reconstructed image is obtained by a rpredefined number of iterations starting with an arbitrary initial image. Its convergence speed depends on the selection of the initial image. It should be solved to get high speed convergence. In this paper, we theoretically reveal that conventional method is approximately decomposed into the decoding of the DC and AC components. Based on this fact, we proposed a novel fast fractal decoding algorithm made up of two steps. The averaged-image considered as an optimal initial image is estimated in the first step. In the second step, the reconstructe dimag eis genrated from the output image obtained in the first step. From the simulations, it is shown that the output image of the first step approximately converges to the averaged-image with only 15% calculations for one iteration of conventional method. And the proposed method is faster than various decoding mehtods and evenly equal to conventioanl decoding with the averaged-image. In addition, the proposed method can be applied to the compressed data resulted from the various encoding methods because it does not impose any constraints in the encoding procedure to get high decoding speed.

• 전기학회논문지
• /
• 제56권6호
• /
• pp.1055-1063
• /
• 2007

This paper introduces an optimal design technique for a 250-watt isolation transformer using an optimization method, dynamic encoding algorithm for searches (DEAS). Although the optimal design technique for transformers dates back to 1970s and various optimization methods have been developed so far, literature concerning global optimization for transformer core design is rarely found against its importance. In this paper, core configuration of the isolation transformer whose performance is computed by complex mathematical steps is optimized with DEAS. The optimization result confirms that DEAS can be successfully employed to transformer core design for various performance specifications only by adjusting weight factors in cost function.

• International Journal of Internet, Broadcasting and Communication
• /
• 제12권2호
• /
• pp.45-50
• /
• 2020

A Language model with neural networks commonly trained with likelihood loss. Such that the model can learn the sequence of human text. State-of-the-art results achieved in various language generation tasks, e.g., text summarization, dialogue response generation, and text generation, by utilizing the language model's next token output probabilities. Monotonous and boring outputs are a well-known problem of this model, yet only a few solutions proposed to address this problem. Several decoding techniques proposed to suppress repetitive tokens. Unlikelihood training approached this problem by penalizing candidate tokens probabilities if the tokens already seen in previous steps. While the method successfully showed a less repetitive generated token, the method has a large memory consumption because of the training need a big vocabulary size. We effectively reduced memory footprint by encoding words as sequences of subword units. Finally, we report competitive results with token level unlikelihood training in several automatic evaluations compared to the previous work.

• 방송과미디어
• /
• 제13권1호
• /
• pp.82-96
• /
• 2008

3차원 오디오 기술은 3차원 오디오를 녹음하는 기술에서 부호화, 전송, 재현에 이르기까지 오디오 신호처리 과정을 포함한다. 하지만 오디오 신호를 재생하는 단말의 환경이 궁극적으로 3차원 오디오를 청취자에게 제공하는 마지막 단계가 되므로 재생환경에 따라서 적용되는 기술에 차이가 있다. 본 논문에서는 음악 콘텐츠를 소비하는 다양한 재생 단말에서 현실감을 증대시키기 위해 적용되는 3차원 오디오 기술의 연구동향을 살펴본다.

• 한국세라믹학회지
• /
• 제56권5호
• /
• pp.421-434
• /
• 2019

We face many fascinating and diverse challenges, the most important among which is to determine how to store a large amount of information with novel approaches. Info-convergence ceramic nanosystems, which combine ceramic materials science and information technology, may provide an attractive alternative. This review considers recent multidisciplinary advances in the development of info-convergence nanosystems based on ceramic materials and discusses various strategies under ceramic-based information systems with a special focus on materials and nanohybridization technologies. Ceramic materials have played diverse roles not only as the generic coding support, but also as the central coding substance. The review highlights the ceramic nanohybrid bio code and ceramic nanoparticle optical code for applications in tracking-and-traceability management, nano-forensics, anti-counterfeiting, and even communication, as well as the four steps of encoding, encrypting, decrypting, and decoding for the desired applications. Additionally, associated challenges, potential solutions, and perspectives for future developments in the field are discussed.

• 전기학회논문지
• /
• 제67권6호
• /
• pp.767-772
• /
• 2018

Recently, some meta-heuristic algorithms, such as GA(Genetic Algorithm) and GP(Genetic Programming), have been used to optimize CNN(Convolutional Neural Network). The CNN, which is one of the deep learning models, has seen much success in a variety of computer vision tasks. However, designing CNN architectures still requires expert knowledge and a lot of trial and error. In this paper, the recent attempts to automatically construct CNN architectures are investigated and analyzed. First, two GA based methods are summarized. One is the optimization of CNN structures with the number and size of filters, connection between consecutive layers, and activation functions of each layer. The other is an new encoding method to represent complex convolutional layers in a fixed-length binary string, Second, CGP(Cartesian Genetic Programming) based method is surveyed for CNN structure optimization with highly functional modules, such as convolutional blocks and tensor concatenation, as the node functions in CGP. The comparison for three approaches is analysed and the outlook for the potential next steps is suggested.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제27권4호
• /
• pp.432-440
• /
• 2000

본 연구에서는 촬상단면 내의 평행이동에 기인한 MRI 아티팩트를 제거하는 새로운 알고리즘으로서, 수렴에 대한 보증이 없는 기존의 반복적인 위상탐색에 의한 알고리즘과는 달리, MRI 촬상원리에 근거한 체동량을 탐색하기 위한 직접적인 방법이 제안된다. 기존의 접근방법에는 x(read out) 방향과 y(phase encoding) 방향의 체동이 동시에 탐색되었다. 하지만 x 방향과 y 방향에 대한 각각의 체동특성이 서로 다르기 때문에 본 방법에서는 이들의 체동특성의 해석에 기초해서 두 단계의 서로 다른 알고리즘을 적용함으로써 x 방향과 y 방향의 각 체동이 제거된다. 우선, x 방향의 체동은 MRI 신호의 x 방향 스펙트럼 변위에 해당하고, 그 스펙트럼의 비제로 영역(non-zero area)은 밀도함수의 x축 투영영역과 일치함에 주목한다. 따라서 체동은 스펙트럼의 비제로 영역과 제로영역(zero area)의 경계를 검출함으로써 추정되며, x 방향의 체동은 역방향으로 스펙트럼을 변위시킴으로써 제거된다. 그 다음으로, y 방향의 체동은 그 체동성분과 실제의 화상성분을 분리할 수 있는 구속조건을 적용함으로써 제거된다. 최종적으로 phantom 화상을 사용한 시뮬레이션을 통해서 본 방법의 유효성을 확인하였으며, 그 결과 제안한 방법이 기존의 방법에 비해 포괄적으로 사용할 수 있음을 알 수 있었다.

• Korean Chemical Engineering Research
• /
• 제57권3호
• /
• pp.321-330
• /
• 2019

최신 학제간 융합기술의 발전과 더불어 디지털 사회에서 생성된 방대한 정보를 저장하기 위하여 신뢰성이 높고 정보의 집적화면에서 효율적이며 세계적으로 통용될 수 있는 정보 코드 시스템의 필요성이 대두되고 있다. 그에 발맞춰 화공, 나노화학, 신소재 분야에서도 획기적인 차세대 정보 코드 시스템 개발을 위한 시도가 계속되고 있다. 해결책 가운데 하나로서 DNA를 이용한 정보 코드 시스템이 각광을 받으면서 많은 연구자들이 관련 연구에 박차를 가하고 있다. 본 총설에서는 DNA를 기반으로 하는 정보 융합 나노하이브리드 (infohybrid) 코드 시스템의 연구동향 및 기술개발 현황에 대하여 요약 및 정리하였다. 특히, 코드 시스템에 적합한 하이브리드 소재 및 나노기술에 대하여 알아보고 대표적인 사례를 기반으로 DNA 정보 코드화 기법에 대하여 정리한 후, 이력추적관리, 진위판별입증, 나노법의학 등 응용분야에 대하여 연구 사례를 중심으로 설명하였다. 마지막으로 최근에 연구 발표된 스마트폰을 활용한 Avatar DNA 나노하이브리드 시스템을 소개함으로써 4차 산업혁명 시대를 이끌어 나갈 나노-바이오-정보-인지 기술이 접목된 융합시스템 개발 및 그 잠재력에 대하여 전망해 보았다.

• 대한의용생체공학회:의공학회지
• /
• 제31권6호
• /
• pp.464-471
• /
• 2010

Compressed sensing can be used to reduce scan time or to enhance spatial resolution in MRI. It is now recognized that compressed sensing works well in reconstructing magnitude images if the sampling mask and the sparsifying transform are well chosen. Phase images also play important roles in MRI particularly in chemical shift imaging and magnetic resonance electrical impedance tomography (MREIT). We reconstruct MRI phase images using the compressed sensing technique. Through computer simulation and real MRI experiments, we reconstructed phase images using the compressed sensing technique and we compared them with the ones reconstructed by conventional Fourier reconstruction technique. As compared to conventional Fourier reconstruction with the same number of phase encoding steps, compressed sensing shows better performance in terms of mean squared phase error and edge preservation. We expect compressed sensing can be used to reduce the scan time or to enhance spatial resolution of MREIT.

• Journal of Multimedia Information System
• /
• 제4권4호
• /
• pp.311-316
• /
• 2017

High Efficiency Video Coding (HEVC) is the newest video coding standard for improvement in video data compression. This new standard provides a significant improvement in picture quality, especially for high-resolution videos. A quadtree-based structure is created for the encoding and decoding processes and the rate-distortion (RD) cost is calculated for all possible dimensions of coding units in the quadtree. To get the best combination of the block an optimization process is performed in the encoder, called rate distortion optimization (RDO). In this work we are proposing a novel approach to enhance the overall RDO process of HEVC encoder. The proposed algorithm is performed in two steps. In the first step, like HEVC, it performs general rate distortion optimization. The second step is an extra checking where a SSIM based cost is evaluated. Moreover, a fast SSIM (FSSIM) calculation technique is also proposed in this paper.