Recently, Kingma & Dhariwal (2018) demonstrated with Glow that generative flows are capable of generating high quality images. We generalize the 1 x 1 convolutions proposed in Glow to invertible d x d convolutions, which are more flexible since they operate on both channel and spatial axes. We propose two methods to produce invertible convolutions, that have receptive fields identical to standard convolutions: Emerging convolutions are obtained by chaining specific autoregressive convolutions, and periodic convolutions are decoupled in the frequency domain. Our experiments show that the flexibility of d x d convolutions significantly improves the performance of generative flow models on galaxy images, CIFAR10 and ImageNet
In this thesis, we study approaches to learn priors on data (i.e. generative modeling) and learners ...
Generative models, especially ones that are parametrized by deep neural networks, are powerful unsup...
136 pagesVisual content is probably the most important medium by which we understand the world. In t...
Treball fi de màster de: Master's Degree in Data Science. Curs 2020-2021Directors: Vicenç Gómez (UPF...
© 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for a...
Flow-based generative models have become an important class of unsupervised learning approaches. In ...
Many measurements or observations in computer vision and machine learning manifest as non-Euclidean ...
Flow-based generative models are an important class of exact inference models that admit efficient i...
Due to the success of generative flows to model data distributions, they have been explored in inver...
Learning a generative model with compositional structure is a fundamental problem in statistics. My ...
Deep Learning is becoming a standard tool across science and industry to optimally solve a variety o...
Generative convolutional deep neural networks, e.g. popular GAN architectures, are relying on convol...
We present progress in developing stable, scalable and transferable generative models for visual dat...
The two key characteristics of a normalizing flow is that it is invertible (in particular, dimension...
While diffusion models have shown great success in image generation, their noise-inverting generativ...
In this thesis, we study approaches to learn priors on data (i.e. generative modeling) and learners ...
Generative models, especially ones that are parametrized by deep neural networks, are powerful unsup...
136 pagesVisual content is probably the most important medium by which we understand the world. In t...
Treball fi de màster de: Master's Degree in Data Science. Curs 2020-2021Directors: Vicenç Gómez (UPF...
© 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for a...
Flow-based generative models have become an important class of unsupervised learning approaches. In ...
Many measurements or observations in computer vision and machine learning manifest as non-Euclidean ...
Flow-based generative models are an important class of exact inference models that admit efficient i...
Due to the success of generative flows to model data distributions, they have been explored in inver...
Learning a generative model with compositional structure is a fundamental problem in statistics. My ...
Deep Learning is becoming a standard tool across science and industry to optimally solve a variety o...
Generative convolutional deep neural networks, e.g. popular GAN architectures, are relying on convol...
We present progress in developing stable, scalable and transferable generative models for visual dat...
The two key characteristics of a normalizing flow is that it is invertible (in particular, dimension...
While diffusion models have shown great success in image generation, their noise-inverting generativ...
In this thesis, we study approaches to learn priors on data (i.e. generative modeling) and learners ...
Generative models, especially ones that are parametrized by deep neural networks, are powerful unsup...
136 pagesVisual content is probably the most important medium by which we understand the world. In t...
.svg%3Fwidth%3D300&w=3840&q=75)
.gif%3Fwidth%3D300&w=3840&q=75)
Add to ORKG