PANet: Few-Shot Image Semantic Segmentation with Prototype Alignment

TL;DR

PANet introduces a prototype alignment network for few-shot image segmentation, leveraging metric learning to generate class-specific prototypes that improve generalization and outperform previous methods on standard benchmarks.

Contribution

The paper proposes a novel prototype alignment network that uses non-parametric metric learning and prototype regularization for improved few-shot segmentation.

Findings

Achieves 48.1% mIoU on PASCAL-5i 1-shot setting.

Surpasses state-of-the-art by 1.8% in 1-shot and 8.6% in 5-shot.

Effectively utilizes support set information for better generalization.

Abstract

Despite the great progress made by deep CNNs in image semantic segmentation, they typically require a large number of densely-annotated images for training and are difficult to generalize to unseen object categories. Few-shot segmentation has thus been developed to learn to perform segmentation from only a few annotated examples. In this paper, we tackle the challenging few-shot segmentation problem from a metric learning perspective and present PANet, a novel prototype alignment network to better utilize the information of the support set. Our PANet learns class-specific prototype representations from a few support images within an embedding space and then performs segmentation over the query images through matching each pixel to the learned prototypes. With non-parametric metric learning, PANet offers high-quality prototypes that are representative for each semantic class and meanwhile discriminative for different classes. Moreover, PANet introduces a prototype alignment regularization between support and query. With this, PANet fully exploits knowledge from the support and provides better generalization on few-shot segmentation. Significantly, our model achieves the mIoU score of 48.1% and 55.7% on PASCAL-5i for 1-shot and 5-shot settings respectively, surpassing the state-of-the-art method by 1.8% and 8.6%.