The Experts below are selected from a list of 11553 Experts worldwide ranked by ideXlab platform

Jing Li - One of the best experts on this subject based on the ideXlab platform.

  • an adaptive scale active contour model for inhomogeneous image segmentation and bias field estimation
    Pattern Recognition, 2018
    Co-Authors: Sanping Zhou, Jing Li
    Abstract:

    Abstract The active contour model is a widely used method for image segmentation. Most existing active contour models yield poor performance when applied to images with severe intensity inhomogeneity. To address this issue, we propose an adaptive-scale active contour model (ASACM) based on image entropy and semi-naive Bayesian classifier, which achieves simultaneous segmentation and bias field estimation for images with severe intensity inhomogeneity. Firstly, an adaptive scale operator is constructed to adaptively adjust the scale of the ASACM according to the degree of the intensity inhomogeneity. Secondly, we define an improved bias field estimation term via distributing a dependent-membership function for each pixel to estimate the bias field in severe inhomogeneous images. Thirdly, a new penalty term is proposed using piecewise polynomial, which helps to avoid time-consuming re-Initialization Process and instability in conventional penalty term. The experimental results demonstrate that the proposed ASACM consistently outperforms many state-of-the-art models in segmentation accuracy, segmentation efficiency and robustness w.r.t Initialization and noise.

Yongsheng Shi - One of the best experts on this subject based on the ideXlab platform.

  • Solving Gravity Anomaly Matching Problem Under Large Initial Errors in Gravity Aided Navigation by Using an Affine Transformation Based Artificial Bee Colony Algorithm
    Frontiers Media S.A., 2019
    Co-Authors: Tian Dai, Lingjuan Miao, Haijun Shao, Yongsheng Shi
    Abstract:

    Gravity aided inertial navigation system (GAINS), which uses earth gravitational anomaly field for navigation, holds strong potential as an underwater navigation system. The gravity matching algorithm is one of the key factors in GAINS. Existing matching algorithms cannot guarantee the matching accuracy in the matching algorithms based gravity aided navigation when the initial errors are large. Evolutionary algorithms, which are mostly have the ability of global optimality and fast convergence, can be used to solve the gravity matching problem under large initial errors. However, simply applying evolutionary algorithms to GAINS may lead to false matching. Therefore, in order to deal with the underwater gravity matching problem, it is necessary to improve the traditional evolutionary algorithms. In this paper, an affine transformation based artificial bee colony (ABC) algorithm, which can greatly improve the positioning precision under large initial errors condition, is developed. The proposed algorithm introduces affine transformation to both Initialization Process and evolutionary Process of ABC algorithm. The single-point matching strategy is replaced by the strategy of matching a sequence of several consecutive position vectors. In addition, several constraints are introduced to the Process of evolution by using the output characteristics of the inertial navigation system (INS). Simulations based on the actual gravity anomaly base map have been performed for the validation of the proposed algorithm

  • Data_Sheet_1_Solving Gravity Anomaly Matching Problem Under Large Initial Errors in Gravity Aided Navigation by Using an Affine Transformation Based Artificial Bee Colony Algorithm.ZIP
    2019
    Co-Authors: Tian Dai, Lingjuan Miao, Haijun Shao, Yongsheng Shi
    Abstract:

    Gravity aided inertial navigation system (GAINS), which uses earth gravitational anomaly field for navigation, holds strong potential as an underwater navigation system. The gravity matching algorithm is one of the key factors in GAINS. Existing matching algorithms cannot guarantee the matching accuracy in the matching algorithms based gravity aided navigation when the initial errors are large. Evolutionary algorithms, which are mostly have the ability of global optimality and fast convergence, can be used to solve the gravity matching problem under large initial errors. However, simply applying evolutionary algorithms to GAINS may lead to false matching. Therefore, in order to deal with the underwater gravity matching problem, it is necessary to improve the traditional evolutionary algorithms. In this paper, an affine transformation based artificial bee colony (ABC) algorithm, which can greatly improve the positioning precision under large initial errors condition, is developed. The proposed algorithm introduces affine transformation to both Initialization Process and evolutionary Process of ABC algorithm. The single-point matching strategy is replaced by the strategy of matching a sequence of several consecutive position vectors. In addition, several constraints are introduced to the Process of evolution by using the output characteristics of the inertial navigation system (INS). Simulations based on the actual gravity anomaly base map have been performed for the validation of the proposed algorithm.

Jungwook Choi - One of the best experts on this subject based on the ideXlab platform.

  • hardware and software co optimization for the Initialization failure of the reram based cross bar array
    ACM Journal on Emerging Technologies in Computing Systems, 2020
    Co-Authors: Youngseok Kim, Vijay Narayanan, Seyoung Kim, Chunchen Yeh, Jungwook Choi
    Abstract:

    Recent advances in deep neural network demand more than millions of parameters to handle and mandate the high-performance computing resources with improved efficiency. The cross-bar array architecture has been considered as one of the promising deep learning architectures that shows a significant computing gain over the conventional Processors. To investigate the feasibility of the architecture, we examine non-idealities and their impact on the performance. Specifically, we study the impact of failed cells due to the Initialization Process of the resistive memory-based cross-bar array. Unlike the conventional memory array, individual memory elements cannot be rerouted and, thus, may have a critical impact on model accuracy. We categorize the possible failures and propose hardware implementation that minimizes catastrophic failures. Such hardware optimization bounds the possible logical value of the failed cells and allows us to compensate for the loss of accuracy via off-line training. By introducing the random weight defects during the training, we show that the model becomes more resilient on the device Initialization failures, therefore, less prone to degrade the inference performance due to the failed devices. Our study sheds light on the hardware and software co-optimization procedure to cope with potentially catastrophic failures in the cross-bar array.

  • hardware and software co optimization for the Initialization failure of the reram based cross bar array
    arXiv: Emerging Technologies, 2020
    Co-Authors: Vijay Narayanan, Jungwook Choi
    Abstract:

    Recent advances in deep neural network demand more than millions of parameters to handle and mandate the high-performance computing resources with improved efficiency. The cross-bar array architecture has been considered as one of the promising deep learning architectures that shows a significant computing gain over the conventional Processors. To investigate the feasibility of the architecture, we examine non-idealities and their impact on the performance. Specifically, we study the impact of failed cells due to the Initialization Process of the resistive memory based cross-bar array. Unlike the conventional memory array, individual memory elements cannot be rerouted and, thus, may have a critical impact on model accuracy. We categorize the possible failures and propose hardware implementation that minimizes catastrophic failures. Such hardware optimization bounds the possible logical value of the failed cells and gives us opportunities to compensate for the loss of accuracy via off-line training. By introducing the random weight defects during the training, we show that the model becomes more resilient on the device Initialization failures, therefore, less prone to degrade the inference performance due to the failed devices. Our study sheds light on the hardware and software co-optimization procedure to cope with potentially catastrophic failures in the cross-bar array.

Songlin Feng - One of the best experts on this subject based on the ideXlab platform.

Alexander J. Flueck - One of the best experts on this subject based on the ideXlab platform.