MATH · IN · MODELS

StyleGAN

NVIDIA

Structures found in this family (6)

Intrinsic-dimension profile across depth, Linear Direction, Linear Separability, Affine Subspace, Linear Subspace, Curvature profile of the representation manifold

By model (5)

StyleGAN2 (trained on FFHQ, 1024x1024)
Intrinsic-dimension profile across depth, Linear Direction, Affine Subspace, Linear Subspace
StyleGAN (trained on FFHQ, 1024x1024)
Linear Direction, Linear Separability
StyleGAN (trained on LSUN Car)
Linear Direction
StyleGAN (trained on LSUN Cat)
Linear Direction
StyleGAN2 (Face256, trained on FFHQ)
Curvature profile of the representation manifold

Observations (8)

Local intrinsic dimension of a real StyleGAN2 latent manifold varies across latent space and correlates with disentanglement
Learned linear latent walks in GANs realize shift, zoom and color edits, with extent tracking training-data variability
GANSpace PCA finds near-independent, unimodal, content-style-separated directions
InterFaceGAN SVM boundary normals are linear, editable, and partially entangled
StyleGAN's intermediate W space scores as more linearly separable (lower separability score) than its input Z space
Top eigenvectors of A^T A of a pretrained GAN generator's first affine layer are linear latent directions that edit semantic attributes
StyleGAN2's W space is globally warped; local bases avoid traversal collapse
Image variation in several real pretrained GANs concentrates along surprisingly few, position-homogeneous major axes of the pullback metric

Papers

Analyzing the Latent Space of GAN through Local Dimension Estimation (2023), On the "steerability" of generative adversarial networks (2020), GANSpace: Discovering Interpretable GAN Controls (2020), InterFaceGAN: Interpreting the Disentangled Face Representation Learned by GANs (2020), A Style-Based Generator Architecture for Generative Adversarial Networks (2018), Closed-Form Factorization of Latent Semantics in GANs (2021), Do Not Escape From the Manifold: Discovering the Local Coordinates on the Latent Space of GANs (2021), The Geometry of Deep Generative Image Models and its Applications (2021)