AdjointBackMapV2: Precise reconstruction of arbitrary CNN unit's activation via adjoint operators. Academic Article

abstract

• Adjoint operators have been found to be effective in the exploration of CNN's inner workings (Wan and Choe, 2022). However, the previous no-bias assumption restricted its generalization. We overcome the restriction via embedding input images into an extended normed space that includes bias in all CNN layers as part of the extended space and propose an adjoint-operator-based algorithm that maps high-level weights back to the extended input space for reconstructing an effective hypersurface. Such hypersurface can be computed for an arbitrary unit in the CNN, and we prove that this reconstructed hypersurface, when multiplied by the original input (through an inner product), will precisely replicate the output value of each unit. We show experimental results based on the CIFAR-10 and CIFAR-100 data sets where the proposed approach achieves near 0 activation value reconstruction error.

authors

• Choe, Yoonsuck

published proceedings

• Neural Netw

author list (cited authors)

• Wan, Q., Cheung, S. W., & Choe, Y.

complete list of authors

• Wan, Qing||Cheung, Siu Wun||Choe, Yoonsuck

publication date

• November 2023

publisher

• Elsevier  Publisher

published in

• Neural Networks  Journal

keywords

• Activation Reconstruction
• Adjoint Operator
• Convolutional Neural Network

PubMed Central ID

• 37977090

Digital Object Identifier (DOI)

• 10.1016/j.neunet.2023.11.009

start page

• 55

end page

• 71

volume

• 170

URL

• http://dx.doi.org/10.1016/j.neunet.2023.11.009