arXiv reaDer
Understanding Neural Network Systems for Image Analysis using Vector Spaces and Inverse Maps
There is strong interest in developing mathematical methods that can be used to understand complex neural networks used in image analysis. In this paper, we introduce techniques from Linear Algebra to model neural network layers as maps between signal spaces. First, we demonstrate how signal spaces can be used to visualize weight spaces and convolutional layer kernels. We also demonstrate how residual vector spaces can be used to further visualize information lost at each layer. Second, we study invertible networks using vector spaces for computing input images that yield specific outputs. We demonstrate our approach on two invertible networks and ResNet18.
updated: Sun Nov 24 2024 10:09:32 GMT+0000 (UTC)
published: Thu Feb 01 2024 01:11:15 GMT+0000 (UTC)
参考文献 (このサイトで利用可能なもの) / References (only if available on this site)
被参照文献 (このサイトで利用可能なものを新しい順に) / Citations (only if available on this site, in order of most recent)
Amazon.co.jpアソシエイト