Towards Fine-grained Image Classification with Generative Adversarial Networks and Facial Landmark Detection

TL;DR

This paper enhances fine-grained image classification by using facial landmark detection and improved GAN-based data augmentation with StyleGAN2-ADA, boosting the performance of Vision Transformer models on the Oxford-IIIT Pets dataset.

Contribution

It introduces a novel approach combining facial landmark detection with advanced GAN augmentation to improve fine-grained classification accuracy.

Findings

GAN-based augmentation improves classification accuracy

Facial landmark cropping enhances image realism

Method outperforms standard augmentation techniques

Abstract

Fine-grained classification remains a challenging task because distinguishing categories needs learning complex and local differences. Diversity in the pose, scale, and position of objects in an image makes the problem even more difficult. Although the recent Vision Transformer models achieve high performance, they need an extensive volume of input data. To encounter this problem, we made the best use of GAN-based data augmentation to generate extra dataset instances. Oxford-IIIT Pets was our dataset of choice for this experiment. It consists of 37 breeds of cats and dogs with variations in scale, poses, and lighting, which intensifies the difficulty of the classification task. Furthermore, we enhanced the performance of the recent Generative Adversarial Network (GAN), StyleGAN2-ADA model to generate more realistic images while preventing overfitting to the training set. We did this by training a customized version of MobileNetV2 to predict animal facial landmarks; then, we cropped images accordingly. Lastly, we combined the synthetic images with the original dataset and compared our proposed method with standard GANs augmentation and no augmentation with different subsets of training data. We validated our work by evaluating the accuracy of fine-grained image classification on the recent Vision Transformer (ViT) Model.