Feature Space Analysis by Guided Diffusion Model

Kimiaki Shirahama, Miki Yanobu, Kaduki Yamashita +1

This paper aims to analyse the feature space of a vision-related Deep Neural Network (DNN) by proposing a decoder that can generate an image whose feature closely matches a user-sp…