Training Convolutional Networks with Web Images

TL;DR

This paper explores using web-collected images to train convolutional neural networks, specifically AlexNet, for large-scale image classification, analyzing different data collection strategies and their impact.

Contribution

It demonstrates the feasibility of building large-scale image datasets from web sources for training deep neural networks, replicating ImageNet with web images.

Findings

Web images can effectively replace traditional datasets for training CNNs.

Different download strategies impact dataset quality and model performance.

Web-based datasets enable scalable and cost-effective training of recognition models.

Abstract

In this thesis we investigate the effect of using web images to build a large scale database to be used along a deep learning method for a classification task. We replicate the ImageNet large scale database (ILSVRC-2012) from images collected from the web using 4 different download strategies varying: the search engine, the query and the image resolution. As a deep learning method, we will choose the Convolutional Neural Network that was very successful with recognition tasks; the AlexNet.