Learning Meta-class Memory for Few-Shot Semantic Segmentation

2021-08-06 · ICCV 2021 10 · Zhonghua Wu, Xiangxi Shi, Guosheng Lin, Jianfei Cai

Currently, the state-of-the-art methods treat few-shot semantic segmentation task as a conditional foreground-background segmentation problem, assuming each class is independent. In this paper, we introduce the concept of meta-class, which is the meta information (e.g. certain middle-level features) shareable among all classes. To explicitly learn meta-class representations in few-shot segmentation task, we propose a novel Meta-class Memory based few-shot segmentation method (MM-Net), where we introduce a set of learnable memory embeddings to memorize the meta-class information during the base class training and transfer to novel classes during the inference stage. Moreover, for the $k$-shot scenario, we propose a novel image quality measurement module to select images from the set of support images. A high-quality class prototype could be obtained with the weighted sum of support image features based on the quality measure. Experiments on both PASCAL-$5^i$ and COCO dataset shows that our proposed method is able to achieve state-of-the-art results in both 1-shot and 5-shot settings. Particularly, our proposed MM-Net achieves 37.5\% mIoU on the COCO dataset in 1-shot setting, which is 5.1\% higher than the previous state-of-the-art.

📄 PDF Abstract BibTeX arXiv:2108.02958

Code (1)

wu-zhonghua/MM-Net pytorch

Tasks

Few-Shot Semantic Segmentation Segmentation Semantic Segmentation

Remember the Difference: Cross-Domain Few-Shot Semantic Segmentation via Meta-Memory Transfer

2022-01-01 · CVPR 2022 1 · Wenjian Wang, Lijuan Duan, Yuxi Wang, Qing En 외

Few-shot semantic segmentation intends to predict pixel level categories using only a few labeled samples. Existing few-shot methods focus primarily on the categories sampled from the same distribution. Nevertheless,…

Contrastive Learning Cross-Domain Few-Shot Few-Shot Learning Few-Shot Semantic Segmentation+2

Differentiable Meta-learning Model for Few-shot Semantic Segmentation

2019-11-23 · Pinzhuo Tian, Zhangkai Wu, Lei Qi, Lei Wang 외

To address the annotation scarcity issue in some cases of semantic segmentation, there have been a few attempts to develop the segmentation model in the few-shot learning paradigm. However, most existing methods only foc…

Few-Shot Learning Few-Shot Semantic Segmentation Meta-Learning Object+2

Multiscale Memory Comparator Transformer for Few-Shot Video Segmentation

2023-07-15 · Mennatullah Siam, Rezaul Karim, He Zhao, Richard Wildes

Few-shot video segmentation is the task of delineating a specific novel class in a query video using few labelled support images. Typical approaches compare support and query features while limiting comparisons to a sing…

Decoder Segmentation Semantic Segmentation Video Object Segmentation+2

Cross-Domain Few-Shot Semantic Segmentation

2022-03-24 · ECCV 2022 3 · Shuo Lei

Few-shot semantic segmentation aims at learning to segment a novel object class with only a few annotated examples. Most existing methods consider a setting where base classes are sampled from the same domain as the no…

Cross-Domain Few-Shot Few-Shot Semantic Segmentation Meta-Learning Segmentation+2

A Pixel-Level Meta-Learner for Weakly Supervised Few-Shot Semantic Segmentation

2021-11-02 · Yuan-Hao Lee, Fu-En Yang, Yu-Chiang Frank Wang

Few-shot semantic segmentation addresses the learning task in which only few images with ground truth pixel-level labels are available for the novel classes of interest. One is typically required to collect a large mount…

Few-Shot Semantic Segmentation Meta-Learning Segmentation Semantic Segmentation