Mobile Computing, Applications, and Services. 10th EAI International Conference, MobiCASE 2019, Hangzhou, China, June 14–15, 2019, Proceedings

Research Article

Progress in Interpretability Research of Convolutional Neural Networks

Download
6 downloads
  • @INPROCEEDINGS{10.1007/978-3-030-28468-8_12,
        author={Wei Zhang and Lizhi Cai and Mingang Chen and Naiqi Wang},
        title={Progress in Interpretability Research of Convolutional Neural Networks},
        proceedings={Mobile Computing, Applications, and Services. 10th EAI International Conference, MobiCASE 2019, Hangzhou, China, June 14--15, 2019, Proceedings},
        proceedings_a={MOBICASE},
        year={2019},
        month={9},
        keywords={Convolutional neural networks “black box” Interpretability},
        doi={10.1007/978-3-030-28468-8_12}
    }
    
  • Wei Zhang
    Lizhi Cai
    Mingang Chen
    Naiqi Wang
    Year: 2019
    Progress in Interpretability Research of Convolutional Neural Networks
    MOBICASE
    Springer
    DOI: 10.1007/978-3-030-28468-8_12
Wei Zhang,*, Lizhi Cai,*, Mingang Chen1,*, Naiqi Wang,*
  • 1: Shanghai Development Center of Computer Software Technology
*Contact email: zweing1995@163.com, clz@ssc.stn.sh.cn, cmg@ssc.stn.sh.cn, slytherinwnq@163.com

Abstract

Convolutional neural networks have made unprecedented breakthroughs in various tasks of computer vision. Due to its complex nonlinear model structure and the high latitude and complexity of data distribution, it has been criticized as an unexplained “black box”. Therefore, explaining the neural network model and uncovering the veil of the neural network have become the focus of attention. This paper starts with the term “interpretability”, summarizes the results of the interpretability of convolutional neural networks in the past three years (2016–2018), and analyses them with interpretable methods. Firstly, the concept of “interpretability” is introduced. Then the existing research achievements are classified and compared from four aspects, data characteristics and rule processing, model internal spatial analysis, interpretation and prediction, and model interpretation. Finally pointed out the possible research directions.