Deep learning and hand-crafted features for virus image classification

TL;DR

This paper combines handcrafted features and deep learning for virus image classification, achieving state-of-the-art accuracy through ensemble methods that leverage both approaches.

Contribution

It introduces a novel ensemble approach that fuses handcrafted and deep learning features for improved virus image classification performance.

Findings

Fusion of features boosts classification accuracy

Deep learning features outperform handcrafted features alone

Ensemble approach achieves state-of-the-art results

Abstract

In this work, we present an ensemble of descriptors for the classification of transmission electron microscopy images of viruses. We propose to combine handcrafted and deep learning approaches for virus image classification. The set of handcrafted is mainly based on Local Binary Pattern variants, for each descriptor a different Support Vector Machine is trained, then the set of classifiers is combined by sum rule. The deep learning approach is a densenet201 pretrained on ImageNet and then tuned in the virus dataset, the net is used as features extractor for feeding another Support Vector Machine, in particular the last average pooling layer is used as feature extractor. Finally, classifiers trained on handcrafted features and classifier trained on deep learning features are combined by sum rule. The proposed fusion strongly boosts the performance obtained by each stand-alone approach, obtaining state of the art performance.