Paper Summaries

MONet: Unsupervised Scene Decomposition and Representation

reimplementation

unsupervised

VAE

The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables

reimplementation

VAE

D-VAE

Importance Weighted Autoencoders

reimplementation

VAE

IWAE

generative

Attend, Infer, Repeat: Fast Scene Understanding with Generative Models

reimplementation

unsupervised

generative

Spatial Transformer Networks

reimplementation

U-Net: Convolutional Networks for Biomedical Image Segmentation

reimplementation

CNN

segmentation

Spatial Broadcast Decoder: A Simple Architecture for Learning Disentangled Representations in VAEs

reimplementation

VAE

disentanglement

Auto-Encoding Variational Bayes

reimplementation

VAE

generative

Schema Networks: Zero-shot Transfer with a Generative Causal Model of Intuitive Physics

RL

Relational Inductive Biases, Deep Learning, and Graph Networks

interaction network

graph networks

generalization

Interaction Networks for Learning about Objects, Relations and Physics

graph networks