Model Stitching and Visualization How GAN Generators can Invert Networks in Real-Time
This provides a fast inversion method for practical applications in domains like digital pathology and image datasets, but it is incremental as it builds on existing GAN and stitching techniques.
The paper tackles the problem of reconstructing activations from classification and segmentation networks by stitching them with a GAN generator using a 1x1 convolution, achieving comparable performance to gradient descent methods but with a processing time two orders of magnitude faster.
In this work, we propose a fast and accurate method to reconstruct activations of classification and semantic segmentation networks by stitching them with a GAN generator utilizing a 1x1 convolution. We test our approach on images of animals from the AFHQ wild dataset, ImageNet1K, and real-world digital pathology scans of stained tissue samples. Our results show comparable performance to established gradient descent methods but with a processing time that is two orders of magnitude faster, making this approach promising for practical applications.