Existing techniques for model inversion typically rely on hard-to-tune regularizers, such as total variation or feature regularization, which must be individually calibrated for each network in order to produce adequate images. In this work, we introduce Plug-In Inversion, which relies on a simple set of augmentations and does not require excessive hyper-parameter tuning. Under our proposed augmentation-based scheme, the same set of augmentation hyper-parameters can be used for inverting a wide range of image classification models, regardless of input dimensions or the architecture. We illustrate the practicality of our approach by inverting Vision Transformers (ViTs) and Multi-Layer Perceptrons (MLPs) trained on the ImageNet dataset, tasks which to the best of our knowledge have not been successfully accomplished by any previous works.
@article{arxiv.2201.12961,
title = {Plug-In Inversion: Model-Agnostic Inversion for Vision with Data Augmentations},
author = {Amin Ghiasi and Hamid Kazemi and Steven Reich and Chen Zhu and Micah Goldblum and Tom Goldstein},
journal= {arXiv preprint arXiv:2201.12961},
year = {2022}
}