Image Retrieval

CSL2050 Course Project 2024

Jay Mehta Akshat Jain Harshiv Shah Jyotin Goel Rhythm Baghel

IIT Jodhpur


Abstract

This project focuses on image retrieval, aiming to retrieve relevant images given an image query. Leveraging both Histogram of Oriented Gradients (HoG) and Convolutional Neural Network (CNN) features extracted through provided implementations, various methodologies are explored, including classification and clustering and mean based techniques. The CIFAR-10 dataset serves as the foundation for this study, offering a diverse collection of images for training and evaluation purposes. Through systematic experimentation and analysis, this project seeks to enhance image retrieval systems, contributing to advancements in image recognition and retrieval technologies..

 

The Image Retrieval Problem



In today's digital age, the vast amount of image data available on the internet poses a significant challenge in efficiently retrieving relevant images based on user queries. To address this challenge, the task at hand is to develop an Image Retrieval System capable of retrieving relevant images given an image query.

 

Youtube Video

 

 

Team

card image

Jay
Mehta

card image

Akshat
Jain

card image

Harshiv Shah

card image

Rhythm
Baghel

card image

Jyotin
Goel

 

 

 

 

Acknowledgment

We are very grateful to Dr Anand Mishra for providing us with this opportunity to work on this project. This project helped us to explore many different techniques for image retrieval which helped us strengthen our basics and learn some of the advanced concepts. We were able to get a hands on experience while working on this project which was very beneficial for all of us.

 

For questions, please contact Jyotin Goel or raise an issue on GitHub.

Copyright © This is an open source project.