Coarse2Fine: Local Consistency Aware Re-prediction for Weakly Supervised Object Localization

Weakly supervised object localization aims to localize objects of interest by using only image-level labels. Existing methods generally segment the activation map by threshold to obtain mask and generate a bounding box. However, the activation map is locally inconsistent, i.e., similar neighboring pixels of the same object are not equally activated, which leads to the blurred boundary issue: the localization result is sensitive to the threshold, and the mask obtained directly from the activation map loses the fine contours of the object, making it difficult to obtain a tight bounding box. In this paper, we introduce the Local Consistency Aware Re-prediction (LCAR) framework, which aims to recover the complete fine object mask from the locally inconsistent activation map and hence obtain a tight bounding box. To this end, we propose the self-guided re-prediction module (SGRM), which employs a novel Aggregation Net with dynamic weights to replace the post-processing of threshold segmentation. To derive more reliable pseudo labels from the activation map to supervise the SGRM, we further design an affinity refinement module (ARM) that utilizes the original image feature to better align the activation map with the image contents and design a selfdistillation CAM (SD-CAM) to alleviate the localizer dependence on saliency. Experiments demonstrate that our LCAR outperforms the state-of-the-art on both the CUB-200-2011 and ILSVRC datasets, achieving 95.9% and 70.7% of GT-Know localization accuracy, respectively.