Improving Fine-Grained Visual Classification using Pairwise Confusion

Fine-Grained Visual Classification (FGVC) datasets contain small sample sizes, along with significant intra-class variation and inter-class similarity. While prior work has addressed intra-class variation using localization and segmentation techniques, the inter-class similarity may also affect feature learning and reduce classification performance. In this work, we address this problem using a novel optimization procedure for the end-to-end neural network training on FGVC tasks. This procedure, called Pairwise Confusion (PC) attempts to learn features with greater generalization, thereby preventing overfitting. This regularization during training is accomplished by intentionally introducing confusion in the activations. With PC regularization, we obtain state-of-the-art performance on six of the most widely-used FGVC datasets and demonstrate improved localization ability. PC is easy to implement, does not need excessive hyperparameter tuning during training, and does not add significant overhead during test time