Image Classification with Classic and Deep Learning Techniques

TL;DR

This paper compares classical computer vision methods and modern deep learning techniques for image classification, evaluating their accuracy and loss across different models including SVM, MLP, InceptionV3, and a custom CNN.

Contribution

It introduces TinyNet, a new CNN architecture designed from scratch, and provides a comparative analysis of classical and deep learning approaches for image classification.

Findings

Accuracy ranges from 0.6 to 0.96 depending on the model.

Deep learning models generally outperform classical methods.

TinyNet shows competitive performance among custom architectures.

Abstract

To classify images based on their content is one of the most studied topics in the field of computer vision. Nowadays, this problem can be addressed using modern techniques such as Convolutional Neural Networks (CNN), but over the years different classical methods have been developed. In this report, we implement an image classifier using both classic computer vision and deep learning techniques. Specifically, we study the performance of a Bag of Visual Words classifier using Support Vector Machines, a Multilayer Perceptron, an existing architecture named InceptionV3 and our own CNN, TinyNet, designed from scratch. We evaluate each of the cases in terms of accuracy and loss, and we obtain results that vary between 0.6 and 0.96 depending on the model and configuration used.