PGNet(Pyramid Graph Networks)[1] modeled structured segmentation data with graphs and further proposed a pyramid-like structure that models different sizes of image regions as graph nodes. There are some details of reading and implementing it.

Contents

• Paper & Code & note
• Paper understanding
• Code understanding
• Note
• References

Paper & Code & note

Paper: Pyramid Graph Networks with Connection Attentions for Region-Based One-Shot Semantic Segmentation(ICCV 2019 paper)
Code: [Code]
Note: Mendeley

Paper

Abstract

1. One-shot image segmentation yields a many-to-many message passing problem with only one training image available.
2. Previous methods described as one-to-many problem by squeezing support data to a global descriptor.
3. In this work, they model structured segmentation data with graphs and apply attentive graph reasoning, graph attention mechanism could establish the element-to-element correspondence, pyramid-like structure is able to capture correspondence at different semantic levels.
4. It leads to new state-of-the-art performance on 1-shot and 5-shot segmentation benchmarks of the PASCAL VOC 2012 dataset.

Problem Description

Problem Solution

Conceptual Understanding

Core Conception

Experiments

Code

[Updating]

Note

[Updating]

References

[1] Zhang C, Lin G, Liu F, et al. Pyramid graph networks with connection attentions for region-based one-shot semantic segmentation[C]//Proceedings of the IEEE International Conference on Computer Vision. 2019: 9587-9595.