ResFeats: Residual Network Based Features for Image Classification

TL;DR

This paper introduces ResFeats, a new image feature extraction method from residual networks that improves classification accuracy across various datasets by leveraging deep residual features and PCA for dimensionality reduction.

Contribution

The paper presents ResFeats, a novel feature extraction approach from residual networks that outperforms traditional CNN features in image classification tasks.

Findings

ResFeats achieve higher accuracy on Caltech-101 and Caltech-256 datasets.

ResFeats significantly improve performance on MIT-67 dataset.

PCA effectively reduces ResFeats dimensionality without losing accuracy.

Abstract

Deep residual networks have recently emerged as the state-of-the-art architecture in image segmentation and object detection. In this paper, we propose new image features (called ResFeats) extracted from the last convolutional layer of deep residual networks pre-trained on ImageNet. We propose to use ResFeats for diverse image classification tasks namely, object classification, scene classification and coral classification and show that ResFeats consistently perform better than their CNN counterparts on these classification tasks. Since the ResFeats are large feature vectors, we propose to use PCA for dimensionality reduction. Experimental results are provided to show the effectiveness of ResFeats with state-of-the-art classification accuracies on Caltech-101, Caltech-256 and MLC datasets and a significant performance improvement on MIT-67 dataset compared to the widely used CNN features.