Image Classification using deep learning network

Project Title

Image Classification using deep learning network

Project Area of Specialization Artificial IntelligenceProject Summary

The purpose of this project is to develop a deep learning classification algorithm which is robust to different lighting conditions and orientations in an image unlike the previous  object classifiers. 

Project Objectives

The following tasks will be performed in this project:

(1)    It will take images from the real world 
(2)    Process the images 
(3)    Make predictions about the category of object in an image 
(4)    Provide results in percentage form of top prediction 

Project Implementation Method

This project is multi-task project which involves the underlying steps:

(a)    First of all the algorithm will be trained on the desired dataset 
(b)    Then the optimal weights of neurons in the network i.e optimal accuracy will be obtained 
(c)    After error reduction and optimization it will be validated on unseen images from the real world to check the performance 
(d)    Then the model will be deployed on the mobile platform 
The software that will be used is Keras with Tensor flow backend. For mobile deployment Android SDK will be used 

Benefits of the Project

This project will help in automation of survillance of areas instead of defining features of detecable objects this algorithm will train on desired dataset and then will predict the object in an image. This project will be another step towards autonomous future. 

Technical Details of Final Deliverable

The expected deliverables of this project are listed below:

(a)    Develop a deep learning algorithm
(b)    Training of deep learning algorithm
(c)    Error reduction and optimization
(d)    Validation of unknown images
(e)    Deployment on mobile platform 
 

Final Deliverable of the Project Hardware SystemType of Industry Security Technologies Artificial Intelligence(AI)Sustainable Development Goals Quality EducationRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	80000
MSI GeForce GTX 1070 Ti 8GB	Equipment	1	70000	70000
MWE Gold 650 Full Modular Power Supply	Miscellaneous	1	10000	10000