A Forward-Backward Approach for Visualizing Information Flow in Deep Networks

A Forward-Backward Approach for Visualizing Information Flow in Deep Networks
复制标题

深度网络中信息流可视化的前向-后向方法

DOI:
--
复制
发表时间:
2017
期刊:
ArXiv
影响因子:
--
通讯作者:
S. Sarkar
S. Sarkar
中科院分区:
--
文献类型:
--
作者:
Aditya Balu;THANH VAN NGUYEN;Apurva Kokate;C. Hegde;S. Sarkar

文献摘要

被引文献

相似文献

我们介绍了一种新的系统框架,用于可视化深度网络中的信息流。具体来说,给定任何训练过的深度卷积网络模型和给定的测试图像,我们的方法在图像域中产生一个紧凑的支持,该支持对应于有助于给定解释的(高分辨率)特征。该方法既具有计算效率,又具有数值鲁棒性。我们提出了几个初步的数值结果,以支持我们的框架优于现有方法。
We introduce a new, systematic framework for visualizing information flow in deep networks. Specifically, given any trained deep convolutional network model and a given test image, our method produces a compact support in the image domain that corresponds to a (high-resolution) feature that contributes to the given explanation. Our method is both computationally efficient as well as numerically robust. We present several preliminary numerical results that support the benefits of our framework over existing methods.