5 papers
Effective Covariance Dynamics in Solvable High-Dimensional GANs
Andrew Bond, Zafer DoÄan
We study a solvable high-dimensional model of generative adversarial network (GAN) training in which a linear generator learns a low-dimensional subspace from data with structured…
Beyond Gaussian Bottlenecks: Topologically Aligned Encoding of Vision-Transformer Feature Spaces
Andrew Bond, Ilkin Umut Melanlioglu, Erkut Erdem +1
Modern visual world modeling systems increasingly rely on high-capacity architectures and large-scale data to produce plausible motion, yet they often fail to preserve underlying 3…
VidStyleODE: Disentangled Video Editing via StyleGAN and NeuralODEs
Moayed Haji Ali, Andrew Bond, Tolga Birdal +4
We propose , a spatiotemporally continuous disentangled eo representation based upon GAN and Neural-s. Effective t…
GaussianVideo: Efficient Video Representation via Hierarchical Gaussian Splatting
Andrew Bond, Jui-Hsien Wang, Long Mai +2
Efficient neural representations for dynamic video scenes are critical for applications ranging from video compression to interactive simulations. Yet, existing methods often face…
Exploring the Precise Dynamics of Single-Layer GAN Models: Leveraging Multi-Feature Discriminators for High-Dimensional Subspace Learning
Andrew Bond, Zafer Dogan
Subspace learning is a critical endeavor in contemporary machine learning, particularly given the vast dimensions of modern datasets. In this study, we delve into the training dyna…