Skip to content

MasazI/cnn_depth_tensorflow

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

cnn_depth_tensorflow

cnn_depth_tensorflow is an implementation of depth estimation using tensorflow.

Original paper is "Depth Map Prediction from a Single Image using a Multi-Scale Deep Network". https://arxiv.org/abs/1406.2283

network

requierments

  • TensorFlow 0.10+
  • Numpy

How to train

  • Download training data. Please see readme.md in data directory.
  • Convert mat to png images.
python prepare_data.py
  • Lets's train.
python task.py
  • You can see predicting images through training in data directory.

example

  • input

- output


Copyright (c) 2016 Masahiro Imai Released under the MIT license

About

depth estimation using tensorflow

Resources

Stars

127 stars

Watchers

10 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages