• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.604-606
• /
• 2024

컴퓨터 비전 응용은 우리 생활에서 중요한 역할을 한다. 현재, 대규모 모델의 등장으로 딥 러닝의 훈련 및 운행 비용이 급격히 상승하고 있다. 자원이 제한된 환경에서는 일부 AI 프로그램을 실행할 수 없게 되므로, 경량화 연구가 필요하다. YOLOv8 은 현재 주요 목표 검출 모델 중 하나이며, 본 논문은 다중 스케일 Ghost 컨볼루션 모듈을 사용하여 구축된 새로운 YOLOv8-pose-msg 키포인트 검출 모델을 제안한다. 다양한 사양에서 새 모델의 매개변수 양은 최소 34% 감소할 수 있으며, 최대 59%까지 감소할 수 있다. 종합적인 검출 성능은 비교적 대규모 데이터셋에서 원래의 수준을 유지할 수 있으며, 소규모 데이터셋에서의 키포인트 검출은 30% 이상 증가할 수 있다. 동시에 최대 25%의 훈련 및 추론 시간을 절약할 수 있다.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.40 no.4
• /
• pp.214-221
• /
• 2003

This paper shows an implementation of video codec (coder/decoder) on an embedded system. The video codec supports both H.261 and H.263 standards. For efficient real-time processing, the video codec is partitioned into a software module and a hardware module. Both modules are codesigned on an embedded system. The software module is processed on a real-time operating system and a RISC processor. It cooperates with the hardware module to compress and decompress images in real time. AMBA (Advanced Microcontroller Bus Architecture) AHB (Advanced High-performance Bus) is used as the system bus. The hardware module works both as AHB masters and as AHB slaves. The encoder part of the hardware module operates in a pipelines mode to compress images in real time. The video codec compresses 15 CIF frames and simultaneously decompresses 15 CIF frames in a second according to H.261 or H.263 standard at 33 MHz frequency.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.12 no.11
• /
• pp.2022-2030
• /
• 2008

System such as FA and breakup TC is applied by real time and need to manage. This paper studies data processing skill that can divide TC and data of much quantity with collection by real time. Modularize to several system, and use of computer communication network that interlink computers that can achieve control function of each systems to network is spreading. Develop that can take advantage of Ideonet communication method and transmit signal of channel because do multiplex all. Do data that have semi-conductor equipment or many input of LCD equipment and output node multiplex, and several units real time Ideonet communication that control is available use that all input of point of contact and output module develop.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.8
• /
• pp.2120-2132
• /
• 1997

A layout automation system for analog integrated circuits is proposed. The implemented system performs full-custom analog layout under the analog layout constraints. In order to overcome the demerits of conventional analog layout systems, parameterized module library is proposed. The system can support complex analog layout modules, resulting in a maximum expandability of the system. Moreover, modified dynamic multi-path algorithm is developed by enhancing the conventional Dijkstra algorithm. Several benchmark circuits such as comparator, op amp, and filter was tested by the system. Layout results compared to OPASYN show well-merging layout and interdigitized layout module.

• The Transactions of the Korea Information Processing Society
• /
• v.5 no.11
• /
• pp.2823-2834
• /
• 1998

This paper deals with the problem of one-to-one mapping of $2^n$ task modules of a parallel program to an n-dimensional hypercube multicomputer so as to minimize to total communication cost during the execution of the task. The problem of finding an optimal mapping has been proven to be NP-complete. We first propose a graph modification technique which transfers the mapping problem in a hypercube multicomputer into the problem of finding a set of maximum cutsets on a given task graph. Using the graph modification technique, we then propose a repeated mapping scheme which efficiently finds a one-to-one mapping of task modules to a hypercube multicomputer by repeatedly applying an existing bipartitioning algorithm on the modified graph. The repeated mapping scheme is shown to be highly effective on a number of test task graphs, it increasingly outperforms the greedy and recursive mapping algorithms as the number of processors increase. The proposed algorithm is shown to be very effective for regular graph, such as hypercube-isomorphic or 'almost' isomorphic graphs and meshes; it finds optimal mapping on almost all the regular task graphs considered.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.1
• /
• pp.67-75
• /
• 1997

This paper presents a virtual-parallel multistage interconnection network (MIN) which provides multipath between processor and memory module. The proposed virtual-parallel MIN network which uses $m{\times}1$ mutiplexer at the input switching block, $1{\times}m$ demultiplexer at the output switching block and logN-1 switching stages has maximum $2{\times}m$ unique paths between processor and memory module. Because it has multi-redundance paths, a number of processors can connect a specific Also, this new virtual-parallel structured MIN network can reduce packet collision possibility at switching block and it has cost. It shown to improve a performance and to be a very simple structure in comparision with MBSF structured MIN.

• Information and Communications Magazine
• /
• v.28 no.11
• /
• pp.16-23
• /
• 2011

휴대폰은 스마트화되면서 퍼스널 컴퓨터에 버금가는 성능으로 발전하고 있다. 또한 모바일 환경에서 높은 데이터 전송 속도 및 많은 사용자가 동시에 사용할 수 있는 시스템을 적용하고 있다. 차세대 이동통신서비스나 B4G(Beyond 4 Generation)에서는 다중밴드/다중 모드/다중경로 처리구조의 융합 단말기가 시장을 형성하게 되며, 갈수록 휴대폰의 RF Front-End는 복잡하게 된다. 또한 FEM(Front End Module)의 성능은 단말기 시스템의 품질(QoS, Quality of Services)을 좌우하는 핵심부품으로 High Data Rate를 위하여 고성능화가 요구되고 있다. 본고에서는 FEM의 발전 트랜드와 구성하는 블록도 및 주요 핵심부품의 특성과 모듈의 소형화 패키징 기술에 관하여 제시하였다. 또한 4G LTE와 Wimax듀얼밴드 Tx/Rx통합 FEM을 소개하였고, 향후에 구성되게 되는 Future FEM의 블록도를 제시하였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10a
• /
• pp.151-153
• /
• 2001

멀티미디어 데이터의 증가와 마이닝 기술의 발전으로 인해 멀티미디어 마이닝에 대한 관심이 증가하고 있다. 본 논문에서는 내용기반의 정보검색 기술과 다차원 다중 데이터큐브 구축기술을 통해 멀터미디어데이타의 마이닝을 구현하는 시스템에 대해 제안한다. 제안 시스템은 멀티미디어 데이터에 내용기반의 정보추출 시스템을 적용하여 성분백터를 추출하고 이를 메타데이타로 한 데이스베이스를 구축한다. 그리고 데이타베이스로부터 지식을 마이닝할 수 있도록 다차원 데이터큐브를 구축하여 빠른 데이터검색과 마이닝결과을 이용자에게 보여주는 모듈로 구성된다. 다차원 데이터큐브는 다중 어레이 구조로써 다차원 데이터를 저장하고, 저장된 여러 데이터 레벨 정보에서 가장 중요한 주제를 통합 생성하여 효율적으로 처리하므로 멀티미디어 데이터를 마이닝하는데 효과적인 방법이다. 또만 다차원데이타큐브를 다중으로 생성하는 방법은 데이터 마이닝 속도를 높이는데 효율적이다.

• KIISE Transactions on Computing Practices
• /
• v.23 no.2
• /
• pp.104-109
• /
• 2017

We propose a new machine learning algorithm which analyzes daily activity patterns of users from multi-modal wearable sensor data. The proposed model learns and extracts activity patterns using input from wearable devices in real-time. Inspired by cue integration of human's property, we constructed gated multi-modal neural networks which integrate wearable sensor input data selectively by using gate modules. For the experiments, sensory data were collected by using multiple wearable devices in restaurant situations. As an experimental result, we first show that the proposed model performs well in terms of prediction accuracy. Then, the possibility to construct a knowledge schema automatically by analyzing the activation patterns in the middle layer of our proposed model is explained.

• Journal of KIISE:Computer Systems and Theory
• /
• v.30 no.3_4
• /
• pp.160-167
• /
• 2003

A static shuffle-exchange network is not only useful for several parallel applications but also use less hardware than the popular multi-stage network or hypercube. Even though it has a lot of advantages, it has never been used in any implemented parallel machine. One of the reasons is there has not been any techniques to make the network fault-tolerant. In this paper multiple fault-tolerant static shuffle-exchange networks are presented. In order to recover from k faulty processing elements, a network needs at least 2 k additional processing elements and at most 4 k additional shuffle ports for each processing elements. By decomposing the k fault-tolerant static shuffle-exchange network into m identical modules, this paper shows that the reliability of the network can be increased.