Densely connected normalizing flows
arXiv:2106.04627
Abstract
Normalizing flows are bijective mappings between inputs and latent representations with a fully factorized distribution. They are very attractive due to exact likelihood valuation and efficient sampling. However, their effective capacity is often insufficient since the bijectivity constraint limits the model width. We address this issue by incrementally padding intermediate representations with noise. We precondition the noise in accordance with previous invertible units, which we describe as cross-unit coupling. Our invertible glow-like modules increase the model expressivity by fusing a densely connected block with Nystrom self-attention. We refer to our architecture as DenseFlow since both cross-unit and intra-module couplings rely on dense connectivity. Experiments show significant improvements due to the proposed contributions and reveal state-of-the-art density estimation under moderate computing budgets.
Accepted at NeurIPS2021
References in corpus (7)
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
- Improved Denoising Diffusion Probabilistic Models
- Flow++: Improving Flow-Based Generative Models with Variational Dequantization and Architecture Design
- Quantifying the Carbon Emissions of Machine Learning
- Hybrid Models with Deep and Invertible Features
- Augmented Normalizing Flows: Bridging the Gap Between Generative Flows and Latent Variable Models