arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1707.09641cs.LGstat.ML
收藏

基于输入恢复的卷积神经网络的可视化建模

Towards Visual Explanations for Convolutional Neural Networks via Input Resampling

Carnegie Mellon University(卡内基梅隆大学);Petuum Inc(Petuum 公司)

Benjamin J. Lengerich, Sandeep Konam, Eric P. Xing, Stephanie Rosenthal, Manuela Veloso

摘要:

神经网络的预测能力往往以模型的可解释性为代价。已经开发了几种技术来解释模型输出的输入功能,但是,很难将这种解释转化为可操作的见解。在这里,我们提出了一个框架,通过检查网络中的信息流来分析模型内部特征的预测。给定经过训练的网络和测试图像,我们通过两个指标选择神经元,这两个指标都是在输入图像扰动创建的一组图像上测量的:(1)神经元激活与网络输出之间相关性的大小和(2)精确度神经元激活。我们发现,前者的度量选择神经元,对网络输出产生很大的影响,而后者的度量选择神经元激活的可推广的功能。通过比较由这两个度量选择的神经元集合,我们的框架提出了一种研究卷积神经网络内部注意力机制的方法。

英文摘要:

The predictive power of neural networks often costs model interpretability. Several techniques have been developed for explaining model outputs in terms of input features; however, it is difficult to translate such interpretations into actionable insight. Here, we propose a framework to analyze predictions in terms of the model's internal features by inspecting information flow through the network. Given a trained network and a test image, we select neurons by two metrics, both measured over a set of images created by perturbations to the input image: (1) magnitude of the correlation between the neuron activation and the network output and (2) precision of the neuron activation. We show that the former metric selects neurons that exert large influence over the network output while the latter metric selects neurons that activate on generalizable features. By comparing the sets of neurons selected by these two metrics, our framework suggests a way to investigate the internal attention mechanisms of convolutional neural networks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑