Liu, W., Wu, Z., Zhao, Y., Fang, Y., Foo, C.-S., Cheng, J., & Lin, G. (2023). Harmonizing Base and Novel Classes: A Class-Contrastive Approach for Generalized Few-Shot Segmentation. International Journal of Computer Vision. https://doi.org/10.1007/s11263-023-01939-y
Abstract:
Current methods for few-shot segmentation (FSSeg) have mainly focused on improving the per- formance of novel classes while neglecting the performance of base classes. To overcome this limitation, the task of generalized few-shot semantic segmentation (GFSSeg) has been intro- duced, aiming to predict segmentation masks for both base and novel classes. However, the current prototype-based methods do not explicitly consider the relationship between base and novel classes when updating prototypes, leading to a limited performance in identifying true categories. To address this challenge, we propose a class contrastive loss and a class relation- ship loss to regulate prototype updates and encourage a large distance between prototypes from different classes, thus distinguishing the classes from each other while maintaining the perfor- mance of the base classes. Our proposed approach achieves new state-of-the-art performance for the generalized few-shot segmentation task on PASCAL VOC and MS COCO datasets.
License type:
Publisher Copyright
Funding Info:
This research / project is supported by the Agency for Science, Technology and Research (A*STAR) - AME Programmatic Funds
Grant Reference no. : A20H6b0151
Description:
This version of the article has been accepted for publication, after peer review and is subject to Springer Nature’s AM terms of use, but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: http://dx.doi.org/10.1007/s11263-023-01939-y