Fine-Grained Image Categorization by Localizing Tiny Object Parts from Unannotated Images

This paper proposes a novel fine-grained image categorization model where no object annotation is required in the training/testing stage. The key technique is a dense graph mining algorithm that localizes multiscale discriminative object parts in each image. In particular, to mimick human hierarchical perception mechanism, a superpixel pyramid is generated for each image, based on which graphlets from each layer are constructed to seamlessly describe object parts. We observe that graphlets representative to each category are densely distributed in the feature space. Therefore a dense graph mining algorithm is developed to discover graphlets representative to each sub-/super-category. Finally, the discovered graphlets from pairwise images are encoded into an image kernel for fine-grained recognition. Experiments on the UCB-200 [32] shown that our method performs competitively to many models relying on the annotated bird parts.