GitHub - ganggit/jointmodel: joint visual restoration and classification

ganggit / jointmodel Public

Notifications You must be signed in to change notification settings
Fork 0
Star 1

joint visual restoration and classification

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commit
JOINT_BACK_CG_4LAYER.m		JOINT_BACK_CG_4LAYER.m
README.txt		README.txt
add_noise.m		add_noise.m
add_noise_sine.m		add_noise_sine.m
dividedbatches.m		dividedbatches.m
evaluate_denoise_classify_4layers.m		evaluate_denoise_classify_4layers.m
getbatches.m		getbatches.m
jointbackprop_dae_4layers.m		jointbackprop_dae_4layers.m
jointlearn_dae_4layers.m		jointlearn_dae_4layers.m
jointlearn_denoise_classification.m		jointlearn_denoise_classification.m
minimize.m		minimize.m
rbm.m		rbm.m
rbmhidlinear.m		rbmhidlinear.m

Repository files navigation

This file provides the joint model for denoising and classification using deep 
4BD8
learning.
It contains the following files:
(1) the two noise models, mentioned in the paper
(2) backpropagation function, JOINT_BACK_CG_4LAYER.m
To make it run, the user shold download the thirdparty toolbox: Autoencoder from Hinton:
http://www.cs.toronto.edu/~hinton/MatlabForSciencePaper.html

For more detail about the model, please refer to:
Gang Chen and Sargur N. Srihari, Joint Visual Restoration and Classification using Deep Learning, in ICDAR2015.