Skip to content

An implementation of the Dynamic Capacity Network paper in Pytorch

Notifications You must be signed in to change notification settings

philqc/Dynamic-Capacity-Network-Pytorch

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 

Repository files navigation

Dynamic-Capacity-Networks

An implementation of the Dynamic Capacity Networks paper in Pytorch (paper at https://arxiv.org/abs/1511.07838)

Based on the tensorflow implementation of the paper found at https://github.com/beopst/dcn.tf.

Used mainly on cluttered-MNIST data set that can be found at https://github.com/deepmind/mnist-cluttered.

Features:

  • Code to compute the receptive field of the coarse and fine model that allow to make quick changes to patch sizes and to the layers of the models.
  • Visualization of the salient patches

Example of visualization of salient patches

Example

To do:

  • Experiment on other data sets
  • Evaluate results for all models and compare them to the paper's results

About

An implementation of the Dynamic Capacity Network paper in Pytorch

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published