Enhancing Image Classification with Augmentation: Data Augmentation Techniques for Improved Image Classification

TL;DR

This study tackles CNN overfitting on small image datasets by introducing three novel data augmentation strategies—pairwise channel transfer, random object occlusion, and novel masking—and evaluating them against standard augmentations on Caltech-101 using EfficientNet-B0. The ensemble of all proposed augmentations yields the best performance, achieving $96.74\%$ accuracy before overfitting, and substantially outperforming existing augmentation methods (approx. $87.78\%$). The results demonstrate that diverse, context-rich augmentations improve generalization and delay overfitting, underscoring the value of combining novel transformations with established techniques. The work suggests broad applicability of these augmentations across domains and motivates further validation on varied datasets.

Abstract

Convolutional Neural Networks (CNNs) serve as the workhorse of deep learning, finding applications in various fields that rely on images. Given sufficient data, they exhibit the capacity to learn a wide range of concepts across diverse settings. However, a notable limitation of CNNs is their susceptibility to overfitting when trained on small datasets. The augmentation of such datasets can significantly enhance CNN performance by introducing additional data points for learning. In this study, we explore the effectiveness of 11 different sets of data augmentation techniques, which include three novel sets proposed in this work. The first set of data augmentation employs pairwise channel transfer, transferring Red, Green, Blue, Hue, and Saturation values from randomly selected images in the database to all images in the dataset. The second set introduces a novel occlusion approach, where objects in the images are occluded by randomly selected objects from the dataset. The third set involves a novel masking approach, using vertical, horizontal, circular, and checkered masks to occlude portions of the images. In addition to these novel techniques, we investigate other existing augmentation methods, including rotation, horizontal and vertical flips, resizing, translation, blur, color jitter, and random erasing, and their effects on accuracy and overfitting. We fine-tune a base EfficientNet-B0 model for each augmentation method and conduct a comparative analysis to showcase their efficacy. For the evaluation and comparison of these augmentation techniques, we utilize the Caltech-101 dataset. The ensemble of image augmentation techniques proposed emerges as the most effective on the Caltech-101 dataset. The results demonstrate that diverse data augmentation techniques present a viable means of enhancing datasets for improved image classification.

Figures (16)

• Figure 1: Schematic diagram of pairwise channel transfer for (HSV) showing the process.
• Figure 2: Schematic diagram showing the Random Object Occlusion process.
• Figure 3: Schematic diagram showing horizontal, vertical, checkered, and circular stripes augmentation.
• Figure 4: Schematic diagram showing various image augmentation techniques.
• Figure 5: Variations of different EfficientNet models. b16