几种深度学习层级化特征学习可视化图的辨析
Hierarchical Representations
Visualization
深度学习(deep learning,DL)的强大得益于其层级化的特征学习能力。常有以下几种图”可视化”DL不同层学习到的特征,以说明模型是如何逐级抽象/提取特征的。
然而一直困惑的是,这几类图有细微差别,甚至矛盾。之前也就人云亦云,今日细究之。
它们分别是
- 卷积核/滤波器的可视化(无监督卷积DBN)
- 反卷积
- 激活最大化(输入空间)
- 特征图
Lee H, Grosse R, Ranganath R, et al. Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations[C]//Proceedings of the 26th annual international conference on machine learning. 2009: 609-616.
Zeiler M D, Fergus R. Visualizing and understanding convolutional networks[C]//European conference on computer vision. Springer, Cham, 2014: 818-833.
Donglai Wei, Bolei Zhou, Antonio Torralba, William T. Freeman(2015): mNeuron: A Matlab Plugin to Visualize Neurons from Deep Models.
https://donglaiw.github.io/proj/mneuron/index.html
https://github.com/donglaiw/mNeuron
https://distill.pub/2017/feature-visualization/
激活最大化其他论文:
Simonyan K, Vedaldi A, Zisserman A. Deep inside convolutional networks: Visualising image classification models and saliency maps[C]//In Workshop at International Conference on Learning Representations. 2014.
Erhan D, Bengio Y, Courville A, et al. Visualizing higher-layer features of a deep network[J]. University of Montreal, 2009, 1341(3): 1.
Taigman Y, Yang M, Ranzato M A, et al. Deepface: Closing the gap to human-level performance in face verification[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2014: 1701-1708.
Original: https://blog.csdn.net/scar_tom/article/details/124240837
Author: scar_tom
Title: 几种深度学习可视化方法(针对层级化特征表示)的辨析
原创文章受到原创版权保护。转载请注明出处:https://www.johngo689.com/690418/
转载文章受原作者版权保护。转载请注明原作者出处!