• 제목/요약/키워드: memory efficiency

검색결과 715건 처리시간 0.027초

Real-time Object Recognition with Pose Initialization for Large-scale Standalone Mobile Augmented Reality

  • Lee, Suwon
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제14권10호
    • /
    • pp.4098-4116
    • /
    • 2020
  • Mobile devices such as smartphones are very attractive targets for augmented reality (AR) services, but their limited resources make it difficult to increase the number of objects to be recognized. When the recognition process is scaled to a large number of objects, it typically requires significant computation time and memory. Therefore, most large-scale mobile AR systems rely on a server to outsource recognition process to a high-performance PC, but this limits the scenarios available in the AR services. As a part of realizing large-scale standalone mobile AR, this paper presents a solution to the problem of accuracy, memory, and speed for large-scale object recognition. To this end, we design our own basic feature and realize spatial locality, selective feature extraction, rough pose estimation, and selective feature matching. Experiments are performed to verify the appropriateness of the proposed method for realizing large-scale standalone mobile AR in terms of efficiency and accuracy.

모바일 메인메모리 데이터베이스 시스템을 위한 효율적인 복구 기법 (An Efficient Recovery Method for Mobile Main Memory Database System)

  • 조성제
    • 한국IT서비스학회지
    • /
    • 제7권2호
    • /
    • pp.181-195
    • /
    • 2008
  • The rapid growth of mobile communication technology has provided the expansion of mobile internet services, particularly mobile realtime transaction takes much weight among mobile fields. There is an increasing demand for various mobile applications to process transactions in a mobile computing fields. Thus, During transmission in wireless networks a base station failure inevitably causes data loss of the base station buffer. It is required to compensate the loss for communication. The existing methods for a base station failure are not adequate because they all suffer from too much overhead and resolve only the link failure. In this paper, we study an efficient recovry systems for a mobile DBMS. We propose SLL (Segment Log List) that enables the mobile host to compensate data loss efficiently in the case of base station failure. In SLL, a base station deliveries an output information of data cells to a mobile host. when a base station fails, the mobile host can retransmit just next data cells. We also prove the efficiency of new method.

The Methodology of Systematic Global Calibration for Process Simulator

  • Lee, Jun-Ha;Lee, Hoong-Joo
    • Transactions on Electrical and Electronic Materials
    • /
    • 제5권5호
    • /
    • pp.180-184
    • /
    • 2004
  • This paper proposes a novel methodology of systematic global calibration and validates its accuracy and efficiency with application to memory and logic devices. With 175 SIMS profiles which cover the range of conditions of implant and diffusion processes in the fabrication lines, the dominant diffusion phenomenon in each process temperature region has been determined. Using the dual-pearson implant model and fully-coupled diffusion model, the calibration was performed systematically. We applied the globally calibrated process simulator parameters to memory and logic devices to predict the optimum process conditions for target device characteristics.

DSP의 DMA 제어기를 이용한 변압기용 디지털 보호계전기의 성능향상 (Efficiency Improvement of Digital Protective Relay for Power Transformer Using DMA Controller of DSP)

  • 권기백;서희석;신명철
    • 대한전기학회논문지:전력기술부문A
    • /
    • 제52권11호
    • /
    • pp.647-654
    • /
    • 2003
  • As electrical power system has become complicated and enlarged to cope with the increasing electric demand, it has to be expected that higher speed, higher reliability, higher function and higher arithmetic ability in protective relay should be realized. Therefore, in this papers, by hardware design and implementation used DMA controller that transfer blocks of data to any location in the memory map without interfering with CPU operation, CPU utilization is increased effectively, as a result it made possible to implement multi-function digital protective relay which has high trust and high function of protection as well as control and metering for power transformers using single processor(DSP).

온칩네트워크를 활용한 DRAM 동시 테스트 기법 (A Concurrent Testing of DRAMs Utilizing On-Chip Networks)

  • 이창진;남종현;안진호
    • 반도체디스플레이기술학회지
    • /
    • 제19권2호
    • /
    • pp.82-87
    • /
    • 2020
  • In this paper, we introduce the novel idea to improve the B/W usage efficiency of on-chip networks used for TAM to test multiple DRAMs. In order to avoid the local bottleneck of test packets caused by an ATE, we make test patterns using microcode-based instructions within ATE and adopt a test bus to transmit test responses from DRAM DFT (Design for Testability) called Test Generator (TG) to ATE. The proposed test platform will contribute to increasing the test economics of memory IC industry.

멀티프로세서 시스템 구성을 위한 시스템 버스의 설계 및 성능평가에 관한 연구 (A Study on the Design and the Performance Evaluation of System Bus for a MC 68000Based Multiprocessor System)

  • 이남재;김영천
    • 한국통신학회논문지
    • /
    • 제15권2호
    • /
    • pp.88-97
    • /
    • 1990
  • 時分割 버스 구조를 갖는 멀티프로세서 시스템의 단점을 보완하기 위하여 각 PE의 로칼메모리를 다른 PE가 직접 액세스하는 것이 가능하도록 시스템 버스를 提案하였다. 이를 위하여 이중 입출 메모리 콘트롤러와 중재기를 설계하였으며 이를 이용한 멀티프로세서 시스템의 具現例를 보였다. 性能評價를 위하여 SPN에 의한 모델링과 부하율에 따른 processing power, 效率 및 시스템 버스의 이용율을 측정하였다.

  • PDF

Photoisomerization of Styrylpyridunium Derivatives for Optical Memory

  • Kang, Young-Soo;Seo, Kyong-Won;Lee, Dong-Jae;Hong, Yong-Pyo
    • KIEE International Transactions on Electrophysics and Applications
    • /
    • 제11C권3호
    • /
    • pp.81-84
    • /
    • 2001
  • The trans and cis forms of N-alkyl-4-styrylpyridinium derivatives (CnSP: n= 4, 8, 12, 16) were successfully synthesized and purified. The derivatives of styrylpyridinium cause photoisomerization when they are illuminated with UV light. The pressure-area isotherms of CnSP and their derivatives were studied to reveal the effect of alkyl chain length. The photoisomerization of CnSP monolayers at the air/water interfaces was indirectly studied by measuring surface tension changes with photoirradiation on the water surface. The characteristics of CnSP were furthermore studied with UV-vis, surface pressure-area isotherms, surface potential-area isotherms, Brewster Angle Microscopy (BAM) at the air/water interface, and optical diffraction efficiency on the ultrathin films.

  • PDF

Magnetization Phenomenon for Variable Flux Memory Motor using Finite Element Method and Experimental Verification

  • Kim, Young Hyun;Lee, Jung Ho
    • Journal of Magnetics
    • /
    • 제21권4호
    • /
    • pp.629-634
    • /
    • 2016
  • In this study, we evaluate the performance of permanent magnets (PMs). The efficiency of attraction in the high speed region was studied using the Variable Flux Memory Motor (VFMM). It is presented in order to analyze the magnetic characteristics of PMs, using the second quadrant plan data with re- and demagnetization. In addition, this study focuses on the evaluation of operational characteristics relative to the magnetizing directions according to the d-axis currents, by using one of the finite element solutions. The feasibility of application for the VFMM has been experimentally demonstrated.

가상화 환경에서 메모리 효율성 향상을 위한 메모리 중복제거 연구 (A Study on Memory Deduplication for Improvement of Memory Efficiency in Virtualization Environment)

  • 조영중;이세호;엄영익
    • 한국정보처리학회:학술대회논문집
    • /
    • 한국정보처리학회 2013년도 추계학술발표대회
    • /
    • pp.21-22
    • /
    • 2013
  • 가상화 기술은 하드웨어 위에서 여러 운영체제를 동작시키면서, 시스템의 활용률을 극대화 시키는 기술이므로 여러 분야에서 각광받고 있다. 가상화는 시스템 위험성 전파 등을 줄임으로써, 보안 노출을 막는 등 여러 장점들이 있다. 하지만, 게스트머신에서 하이퍼바이저로의 잦은 스위치는 가상화 성능을 떨어트린다. 또한, 다수의 가상머신에서 공유될 수 있는 페이지들에 대한 메모리 중복 문제도 존재한다. 우리는 가상화 환경에서 VMEXIT를 줄이고, 메모리를 절약할 수 있는 메모리 중복제거 기술을 제안하고, 이를 정성적으로 성능 평가를 진행하였다.

A hybrid deep learning model for predicting the residual displacement spectra under near-fault ground motions

  • Mingkang Wei;Chenghao Song;Xiaobin Hu
    • Earthquakes and Structures
    • /
    • 제25권1호
    • /
    • pp.15-26
    • /
    • 2023
  • It is of great importance to assess the residual displacement demand in the performance-based seismic design. In this paper, a hybrid deep learning model for predicting the residual displacement spectra under near-fault (NF) ground motions is proposed by combining the long short-term memory network (LSTM) and back-propagation (BP) network. The model is featured by its capacity of predicting the residual displacement spectrum under a given NF ground motion while considering the effects of structural parameters. To construct this model, 315 natural and artificial NF ground motions were employed to compute the residual displacement spectra through elastoplastic time history analysis considering different structural parameters. Based on the resulted dataset with a total of 9,450 samples, the proposed model was finally trained and tested. The results show that the proposed model has a satisfactory accuracy as well as a high efficiency in predicting residual displacement spectra under given NF ground motions while considering the impacts of structural parameters.