.png){kind=icon}
Return
Weakly-supervised Incremental learning for Semantic segmentation with Class Hierarchy
.png){kind=icon}
PRE
AI
DOI：10.1016/j.patrec.2024.04.006.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Although current semantic segmentation approaches have achieved impressive performance, their ability to incrementally learn new classes is limited. Moreover, pixel -by -pixel annotations are costly and time-consuming. Therefore, a new field called Weakly-supervised Incremental Learning for Semantic Segmentation (WILSS) has emerged, which learns new classes using image-level labels. However, image-level labels do not provide sufficient detail, and we discover that the state -of -the -art of WILSS suffers from confusion between old knowledge and new knowledge. To address this issue, we propose W eakly-supervised I ncremental learning for S emantic segmentation with Class H ierarchy (WISH), a method that considers the hierarchical structure of each class when determining which knowledge to trust in cases of confusion between old and new knowledge. Our method has achieved new state -of -the -art performances in all settings compared to the previous methods on the Pascal VOC and MS COCO datasets.
Keywords:
Machine learning
Deep learning
Semantic segmentation
Weak supervision
Incremental learning
Journal
IF:
3.3
Papers:
8.0K
Citations:
1.6W
Organization
Cited Papers
ShareSaveSYNOVIAL FLUID POLYMORPHONUCLEAR LEUCOCYTES FROM PATIENTS WITH RHEUMATOID ARTHRITIS HAVE REDUCED MPO AND NADPH-OXIDASE ACTIVITY.png)
Rheumatology
RheumatologyIF0
PREAI
ShareSaveShareSaveShareSaveShareSaveShareSaveAcoustic vector sensor analysis of the Monterey Bay region soundscape and the impact of COVID-19
OAAI
OAAIShareSaveShareSaveShareSave
View more