2021/08/17 by Weide Liu, Xiangfei Kong, Liu, Weide +5 · 2 citations
Computer Science · #Advanced Image and Video Retrieval Techniques #Advanced Neural Network Applications #Artificial intelligence #Computer science #Computer vision #Generality #Image (mathematics) #Image segmentation #Medical Image Segmentation Techniques #Object (grammar) #Pascal (unit) #Pattern recognition (psychology) #Robustness (evolution) #Segmentation #cs.CV
paper · pdf · doi:10.48550/arxiv.2108.07413
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2021/08/17 · arxiv created 2022/06/29 · arxiv updated 2022/06/30 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/06
Weakly supervised image segmentation trained with image-level labels usually suffers from inaccurate coverage of object areas during the generation of the pseudo groundtruth. This is because the object activation maps are trained with the classification objective and lack the ability to generalize. To improve the generality of the objective activation maps, we propose a region prototypical network RPNet to explore the cross-image object diversity of the training set. Similar object parts across images are identified via region feature comparison. Object confidence is propagated between regions to discover new object areas while background regions are suppressed. Experiments show that the proposed method generates more complete and accurate pseudo object masks, while achieving state-of-the-art performance on PASCAL VOC 2012 and MS COCO. In addition, we investigate the robustness of the proposed method on reduced training sets.