Learning to Navigate for Fine-grained Classification

Fine-grained classification is challenging due to the difficulty of finding discriminative features. Finding those subtle traits that fully characterize the object is not straightforward. To handle this circumstance, we propose a novel self-supervision mechanism to effectively localize informative regions without the need of bounding-box/part annotations. Our model, termed NTS-Net for Navigator-Teacher-Scrutinizer Network, consists of a Navigator agent, a Teacher agent and a Scrutinizer agent. In consideration of intrinsic consistency between informativeness of the regions and their probability being ground-truth class, we design a novel training paradigm, which enables Navigator to detect most informative regions under the guidance from Teacher. After that, the Scrutinizer scrutinizes the proposed regions from Navigator and makes predictions. Our model can be viewed as a multi-agent cooperation, wherein agents benefit from each other, and make progress together. NTS-Net can be trained end-to-end, while provides accurate fine-grained classification predictions as well as highly informative regions during inference. We achieve state-of-the-art performance in extensive benchmark datasets.

PDF Abstract ECCV 2018 PDF ECCV 2018 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Fine-Grained Image Classification CUB-200-2011 NTS-Net (K = 4) Accuracy 87.5% # 54
Fine-Grained Image Classification FGVC Aircraft NTS-Net (K=4) Accuracy 91.4% # 42
Fine-Grained Image Classification Stanford Cars NTS-Net (K=4) Accuracy 93.9% # 50

Methods


No methods listed for this paper. Add relevant methods here