Closed-Form Factorization of Latent Semantics in GANs
Authors:
Top Cited Papers
- 1 June 2021
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1532-1540
Abstract
A rich set of interpretable dimensions has been shown to emerge in the latent space of the Generative Adversarial Networks (GANs) trained for synthesizing images. In order to identify such latent dimensions for image editing, previous methods typically annotate a collection of synthesized samples and train linear classifiers in the latent space. However, they require a clear definition of the target attribute as well as the corresponding manual annotations, limiting their applications in practice. In this work, we examine the internal representation learned by GANs to reveal the underlying variation factors in an unsupervised manner. In particular, we take a closer look into the generation mechanism of GANs and further propose a closedform factorization algorithm for latent semantic discovery by directly decomposing the pre-trained weights. With a lightning-fast implementation, our approach is capable of not only finding semantically meaningful dimensions comparably to the state-of-the-art supervised methods, but also resulting in far more versatile concepts across multiple GAN models trained on a wide range of datasets. 1Keywords
All Related Versions
This publication has 10 references indexed in Scilit:
- InterFaceGAN: Interpreting the Disentangled Face Representation Learned by GANsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2020
- Image Processing Using Multi-Code GAN Prior2020
- Interpreting the Latent Space of GANs for Semantic Face Editing2020
- GANalyze: Toward Visual Definitions of Cognitive Image Properties2019
- A Style-Based Generator Architecture for Generative Adversarial Networks2019
- Arbitrary Style Transfer in Real-Time with Adaptive Instance Normalization2017
- Deep Residual Learning for Image Recognition2016
- Deep Learning Face Attributes in the Wild2015
- Streetscore -- Predicting the Perceived Safety of One Million Streetscapes2014
- ImageNet: A large-scale hierarchical image database2009