Finding Original Image Of A Sub Image Using CNNs

TL;DR

This paper introduces a CNN-based method to identify whether a test image is part of a larger original image, offering an efficient, unsupervised alternative to traditional image similarity and object detection techniques.

Contribution

The paper presents a novel CNN approach for sub-image identification that requires fewer examples, is unsupervised, and more efficient than existing methods like K-Means or K-NN.

Findings

The CNN-based method accurately identifies sub-images with high precision.

It outperforms traditional unsupervised algorithms in efficiency and accuracy.

The approach requires fewer training examples and less computational time.

Abstract

Convolututional Neural Networks have achieved state of the art in image classification, object detection and other image related tasks. In this paper I present another use of CNNs i.e. if given a set of images and then giving a single test image the network identifies that the test image is part of which image from the images given before. This is a task somehow similar to measuring image similarity and can be done using a simple CNN. Doing this task manually by looping can be quite a time consuming problem and won't be a generalizable solution. The task is quite similar to doing object detection but for that lots training data should be given or in the case of sliding window it takes lot of time and my algorithm can work with much fewer examples, is totally unsupervised and works much efficiently. Also, I explain that how unsupervised algorithm like K-Means or supervised algorithm like K-NN are not good enough to perform this task. The basic idea is that image encodings are collected for each image from a CNN, when a test image comes it is replaced by a part of original image, the encoding is generated using the same network, the frobenius norm is calculated and if it comes under a tolerance level then the test image is said to be the part of the original image.