Self-regularized prototypical network for few-shot semantic segmentation

作者：

Highlights：

• We propose a direct yet effective self-regularization module. Prototypes are generated, evaluated, and regularized under the supervision of support masks, which differs from existing works.

• We adopt fidelity as the distance metric in prototype generation for the first time, which provides a more evident distinction for vectors.

• We adopt an iterative query inference module, which uses a collection of prototypes for segmentation and improves the generalization ability for query inference.

• We achieve new state-of-the-art performance on two few-shot segmentation benchmarks.

摘要

•We propose a direct yet effective self-regularization module. Prototypes are generated, evaluated, and regularized under the supervision of support masks, which differs from existing works.•We adopt fidelity as the distance metric in prototype generation for the first time, which provides a more evident distinction for vectors.•We adopt an iterative query inference module, which uses a collection of prototypes for segmentation and improves the generalization ability for query inference.•We achieve new state-of-the-art performance on two few-shot segmentation benchmarks.

论文关键词：Few-shot segmentation,Prototype,Prototypical network,Self-regularized,Non-parametric distance fidelity,Iterative query inference,SRPNet,CNN

论文评审过程：Received 27 April 2022, Revised 10 August 2022, Accepted 4 September 2022, Available online 6 September 2022, Version of Record 13 September 2022.

论文官网地址：https://doi.org/10.1016/j.patcog.2022.109018