vae_vpflows

Flow library

A PyTorch implementation of volume-preserving flows for variational autoencoders.

Code in PyTorch for the convex combination linear IAF and the Householder Flow, J.M. Tomczak & M. Welling

GitHub

90 stars
2 watching
15 forks
Language: Python
last commit: over 7 years ago

Related projects:

Repository Description Stars
kefirski/pytorch_rvae A deep learning implementation of a recurrent variational autoencoder for generating sequential data. 357
ikostrikov/pytorch-flows PyTorch implementations of algorithms for density estimation using invertible transformations. 577
anuragranj/back2future.pytorch An implementation of unsupervised learning for multi-frame optical flow with occlusions using PyTorch. 112
vinhkhuc/pytorch-mini-tutorials A collection of tutorials and lessons on building deep learning models using the PyTorch library. 326
serycjon/vpype-flow-imager Generates flowline representation from an image. 168
schlumberger/joint-vae A PyTorch implementation of Learning Disentangled Joint Continuous and Discrete Representations 462
localdevices/pyorc An image-based river flow analysis library 145
fartashf/vsepp A PyTorch implementation of visual-semantic embedding methods for image-caption retrieval 492
jmhessel/fmpytorch PyTorch implementation of a Factorization Machine module 173
mattmacy/vnet.pytorch A PyTorch implementation of V-Net for volumetric medical image segmentation 703
wmvanvliet/pytorch_hmax An implementation of the HMAX model for vision processing in PyTorch 57
cadene/vqa.pytorch A PyTorch implementation of visual question answering with multimodal representation learning 718
clementpinard/flownetpytorch PyTorch implementation of a neural network architecture for estimating the optical flow between frames in videos 861
openseg-group/openseg.pytorch Provides a PyTorch implementation of several computer vision tasks including object detection, segmentation and parsing. 1,191
zuoxingdong/vin_pytorch_visdom Implementation of Value Iteration Networks in PyTorch with visualization capabilities using Visdom. 226