Interpreting Spatially Infinite Generative Models

Traditional deep generative models of images and other spatial modalities can only generate fixed sized outputs. The generated images have exactly the same resolution as the training images, which is dictated by the number of layers in the underlying neural network. Recent work has shown, however, that feeding spatial noise vectors into a fully convolutional neural network enables both generation of arbitrary resolution output images as well as training on arbitrary resolution training images. While this work has provided impressive empirical results, little theoretical interpretation was provided to explain the underlying generative process. In this paper we provide a firm theoretical interpretation for infinite spatial generation, by drawing connections to spatial stochastic processes. We use the resulting intuition to improve upon existing spatially infinite generative models to enable more efficient training through a model that we call an infinite generative adversarial network, or $\infty$-GAN. Experiments on world map generation, panoramic images and texture synthesis verify the ability of $\infty$-GAN to efficiently generate images of arbitrary size.

[1]  Sepp Hochreiter,et al.  GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium , 2017, NIPS.

[2]  P. Robinson,et al.  The estimation of a nonlinear moving average model , 1977 .

[3]  Nal Kalchbrenner,et al.  Generating High Fidelity Images with Subscale Pixel Networks and Multidimensional Upscaling , 2018, ICLR.

[4]  Daisuke Komura,et al.  Machine Learning Methods for Histopathological Image Analysis , 2017, Computational and structural biotechnology journal.

[5]  Xiaogang Wang,et al.  StackGAN++: Realistic Image Synthesis with Stacked Generative Adversarial Networks , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[6]  Marc Levoy,et al.  Fast texture synthesis using tree-structured vector quantization , 2000, SIGGRAPH.

[7]  J. Olsen Realtime Procedural Terrain Generation , 2004 .

[8]  J. S. Huang,et al.  THE AUTOREGRESSIVE MOVING AVERAGE MODEL FOR SPATIAL ANALYSIS , 1984 .

[9]  Ken Perlin,et al.  An image synthesizer , 1988 .

[10]  Timo Aila,et al.  A Style-Based Generator Architecture for Generative Adversarial Networks , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[11]  Soumith Chintala,et al.  Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks , 2015, ICLR.

[12]  Jan Kautz,et al.  High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[13]  Roland Vollgraf,et al.  Learning Texture Manifolds with the Periodic Spatial GAN , 2017, ICML.

[14]  Christian Ledig,et al.  Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[15]  Tali Dekel,et al.  SinGAN: Learning a Generative Model From a Single Natural Image , 2019, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[16]  Leon A. Gatys,et al.  A Neural Algorithm of Artistic Style , 2015, ArXiv.

[17]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[18]  Andrea Vedaldi,et al.  Improved Texture Networks: Maximizing Quality and Diversity in Feed-Forward Stylization and Texture Synthesis , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[19]  F. Kenton Musgrave,et al.  The synthesis and rendering of eroded fractal terrains , 1989, SIGGRAPH.

[20]  Alexei A. Efros,et al.  Image quilting for texture synthesis and transfer , 2001, SIGGRAPH.

[21]  Alexei A. Efros,et al.  Image-to-Image Translation with Conditional Adversarial Networks , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[22]  R. Haining The moving average model for spatial interaction , 1978 .

[23]  L. Anselin Spatial Econometrics: Methods and Models , 1988 .

[24]  Jeff Donahue,et al.  Large Scale GAN Training for High Fidelity Natural Image Synthesis , 2018, ICLR.

[25]  Roland Vollgraf,et al.  Texture Synthesis with Spatial Generative Adversarial Networks , 2016, ArXiv.

[26]  Christopher Joseph Pal,et al.  A step towards procedural terrain generation with GANs , 2017, ArXiv.

[27]  Leon A. Gatys,et al.  Texture Synthesis Using Convolutional Neural Networks , 2015, NIPS.

[28]  Sergio Gomez Colmenarejo,et al.  Parallel Multiscale Autoregressive Density Estimation , 2017, ICML.

[29]  Rob Fergus,et al.  Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks , 2015, NIPS.

[30]  Irfan A. Essa,et al.  Graphcut textures: image and video synthesis using graph cuts , 2003, ACM Trans. Graph..

[31]  Andrea Vedaldi,et al.  Instance Normalization: The Missing Ingredient for Fast Stylization , 2016, ArXiv.

[32]  Lucas Theis,et al.  Amortised MAP Inference for Image Super-resolution , 2016, ICLR.

[33]  Prafulla Dhariwal,et al.  Glow: Generative Flow with Invertible 1x1 Convolutions , 2018, NeurIPS.

[34]  Andrea Vedaldi,et al.  Texture Networks: Feed-forward Synthesis of Textures and Stylized Images , 2016, ICML.

[35]  Tieniu Tan,et al.  IntroVAE: Introspective Variational Autoencoders for Photographic Image Synthesis , 2018, NeurIPS.

[36]  Aaron C. Courville,et al.  Improved Training of Wasserstein GANs , 2017, NIPS.

[37]  Li Fei-Fei,et al.  Perceptual Losses for Real-Time Style Transfer and Super-Resolution , 2016, ECCV.

[38]  Donald S. Fussell,et al.  Computer rendering of stochastic models , 1982, Commun. ACM.

[39]  Yoshua Bengio,et al.  Generative Adversarial Nets , 2014, NIPS.

[40]  Adam Finkelstein,et al.  PatchMatch: a randomized correspondence algorithm for structural image editing , 2009, SIGGRAPH 2009.

[41]  Chuan Li,et al.  Precomputed Real-Time Texture Synthesis with Markovian Generative Adversarial Networks , 2016, ECCV.

[42]  Ken Perlin,et al.  Improving noise , 2002, SIGGRAPH.

[43]  José Miguel Hernández-Lobato,et al.  Variational Implicit Processes , 2018, ICML.

[44]  Jaakko Lehtinen,et al.  Progressive Growing of GANs for Improved Quality, Stability, and Variation , 2017, ICLR.

[45]  Dimitris N. Metaxas,et al.  StackGAN: Text to Photo-Realistic Image Synthesis with Stacked Generative Adversarial Networks , 2016, 2017 IEEE International Conference on Computer Vision (ICCV).