RG-Flow: A hierarchical and explainable flow model based on renormalization group and sparse prior

Abstract

Flow-based generative models have become an important class of unsupervised learning approaches. In this work, we incorporate the key ideas of renormalization group (RG) and sparse prior distribution to design a hierarchical flow-based generative model, RG-Flow, which can separate information at different scales of images and extract disentangled representations at each scale. We demonstrate our method on synthetic multi-scale image datasets and the CelebA dataset, showing that the disentangled representations enable semantic manipulation and style mixing of the images at different scales. To visualize the latent representations, we introduce receptive fields for flow-based models and show that the receptive fields of RG-Flow are similar to those of convolutional neural networks. In addition, we replace the widely adopted isotropic Gaussian prior distribution by the sparse Laplacian distribution to further enhance the disentanglement of representations. From a theoretical perspective, our proposed method has $O(\log L)$ complexity for inpainting of an image with edge length $L$, compared to previous generative models with $O(L^2)$ complexity.

31 pages, 20 figures, 3 tables

References in corpus (15)

• NICE: Non-linear Independent Components Estimation
• A class of quantum many-body states that can be efficiently simulated
• NVAE: A Deep Hierarchical Variational Autoencoder
• An exact mapping between the Variational Renormalization Group and Deep Learning
• Discovering Hidden Factors of Variation in Deep Networks
• Normalizing Flows on Tori and Spheres
• AdS/CFT as a deep Boltzmann machine
• Deep Learning and AdS/QCD
• Holography as deep learning
• Emerging Convolutions for Generative Normalizing Flows
• Flows for simultaneous manifold learning and density estimation
• Understanding Anomaly Detection with Deep Invertible Networks through Hierarchies of Distributions and Features
• Word Embedding Visualization Via Dictionary Learning
• Transflow Learning: Repurposing Flow Models Without Retraining
• Criticality & Deep Learning II: Momentum Renormalisation Group

Cited by in corpus (3)

• Neural Information Squeezer for Causal Emergence
• Differentiable Programming of Isometric Tensor Networks
• Learning Non-linear Wavelet Transformation via Normalizing Flow