【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun

   【深度学习】Backpropagation applied to Handwritten zip code recognition ----LeCun

    此篇注解LeCun的CNN名作,主要讲述了反向传播算法在手写邮政编码识别中的应用

    作者:小萨

    转载请注明出处:https://blog.csdn.net/u012679707/article/details/80738633


一.整篇结构

    1.introduction
    2.zip codes
          2.1 data base
          2.2 preprocessing
     3.Network Design
          3.1 input and output
          3.2 Feature map and Weight sharing
          3.3 Network architecture
    4.Experiment environment
    5.Results
          5.1 comparision with other work
          5.2 DSP implementation

    6.Conclusion

二.文章提出的深度网络结构

                【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第1张图片


【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第2张图片


三.文章注解

      【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第3张图片

         【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第4张图片

            【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第5张图片

                【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第6张图片

                    【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第7张图片

                        【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第8张图片

                            【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第9张图片


                            【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第10张图片


                                   【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第11张图片

                                    【深度学习经典论文】Backpropagation applied to Handwritten zip code recognition ----LeCun_第12张图片

                    -------------------------------------------         END      -------------------------------------

你可能感兴趣的:(深度学习)