Journal article Open Access

Fine-Grained Adversarial Semi-supervised Learning

Daniele Mugnai; Federico Pernici; Francesco Turchini; Alberto del Bimbo

In this paper we exploit Semi-Supervised Learning (SSL) to increase the amount of training data to improve the performance of
Fine-Grained Visual Categorization (FGVC). This problem has not been investigated in the past in spite of prohibitive annotation costs that FGVC requires. Our approach leverages unlabeled data with an adversarial optimization strategy in which the internal features representation is obtained with a second-order pooling model. This combination allows to back-propagate the information of the parts, represented by second-order pooling, onto unlabeled data in an adversarial training setting. We demonstrate the effectiveness of the combined use by conducting experiments on six state-of-the-art fi ne-grained datasets, which include Aircrafts, Stanford Cars, CUB-200-2011, Oxford Flowers, Stanford Dogs, and the recent Semi-Supervised iNaturalist-Aves. Experimental results clearly show that our proposed method has better performance than the only previous approach that examined this problem; it also obtained higher classification accuracy with respect to the supervised learning methods with which we compared.

Files (3.1 MB)
Name Size
TOMM-adversarial-fine-grained.pdf
md5:8cb25cc8039960b7e30a878d2878c260
3.1 MB Download
65
33
views
downloads
All versions This version
Views 6565
Downloads 3333
Data volume 100.8 MB100.8 MB
Unique views 3737
Unique downloads 3131

Share

Cite as