Boosting Few-shot Fine-grained Recognition with Background Suppression and Foreground Alignment

4 Oct 2022  ·  Zican Zha, Hao Tang, Yunlian Sun, Jinhui Tang ·

Few-shot fine-grained recognition (FS-FGR) aims to recognize novel fine-grained categories with the help of limited available samples. Undoubtedly, this task inherits the main challenges from both few-shot learning and fine-grained recognition. First, the lack of labeled samples makes the learned model easy to overfit. Second, it also suffers from high intra-class variance and low inter-class differences in the datasets. To address this challenging task, we propose a two-stage background suppression and foreground alignment framework, which is composed of a background activation suppression (BAS) module, a foreground object alignment (FOA) module, and a local-to-local (L2L) similarity metric. Specifically, the BAS is introduced to generate a foreground mask for localization to weaken background disturbance and enhance dominative foreground objects. The FOA then reconstructs the feature map of each support sample according to its correction to the query ones, which addresses the problem of misalignment between support-query image pairs. To enable the proposed method to have the ability to capture subtle differences in confused samples, we present a novel L2L similarity metric to further measure the local similarity between a pair of aligned spatial features in the embedding space. What's more, considering that background interference brings poor robustness, we infer the pairwise similarity of feature maps using both the raw image and the refined image. Extensive experiments conducted on multiple popular fine-grained benchmarks demonstrate that our method outperforms the existing state of the art by a large margin. The source codes are available at: https://github.com/CSer-Tang-hao/BSFA-FSFG.

PDF Abstract

Results from the Paper


  Submit results from this paper to get state-of-the-art GitHub badges and help the community compare results to other papers.

Methods


No methods listed for this paper. Add relevant methods here