Skip to content

This repository will guide in applying Deep Learning in Vision. It contains all the work I did in Tensorflow while learning.

Notifications You must be signed in to change notification settings

KillerStrike17/Deep-Learning-for-Vision-Using-Tensorflow-Keras

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Deep Learning in Vision using Tensorflow & Keras

This repository is about applied Deep Learning concepts in the domain of vision from the basics.

This will give a brief idea on what is a kernel and channels, what is its job and how to write custom kernels.

This will lead to understand and calculate local and global receptive fields.

This contains vanilla code to develop your first Deep Neural Network Model over MNIST dataset. The notebook shared also contains the code to view the applied filters over to different layers.

This will guide you on how to improve your vanilla network over accuracy achieved and the size of model. This will also guide in what architecture to use.

Reduce overfitting by L2 regularization Techniques. The folder also has a code to obtain misclassified images, which will help in understanding where the model is failing.

There are various types of Convolution layers available, The Notebooks will contains its implementation on how to use what and when not to use what. This notebook is just to show how to integrate it with your model.

This module will consist of calculating receptive field for GoogleNet, which will help in clearing all your doubts regarding receptive fields. Then we have tried to replicate the SOTA ENAS Network.

About

This repository will guide in applying Deep Learning in Vision. It contains all the work I did in Tensorflow while learning.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published