Objectness-Aware Few-Shot Semantic Segmentation
Preprint
Abstract
Few-shot semantic segmentation models aim to segment images after learning from only a few annotated examples. A key challenge for them is overfitting. Prior works usually limit the overall model capacity to alleviate overfitting, but the limited capacity also hampers the segmentation accuracy. We instead propose a method that increases the overall model capacity by supplementing class-specific features with objectness, which is class-agnostic and so not prone to overfitting. Extensive experiments demonstrate the versatility of our method with multiple backbone models (ResNet-50, ResNet-101 and HRNetV2-W48) and existing base architectures (DENet and PFENet). Given only one annotated example of an unseen category, experiments show that our method outperforms state-of-art methods with respect to mIoU by at least 4.7% and 1.5% on PASCAL-5i and COCO-20i respectively.Keywords
All Related Versions
This publication has 13 references indexed in Scilit:
- Few-Shot Semantic Segmentation with Democratic Attention Networks2020
- Prototype Mixture Models for Few-Shot Semantic Segmentation2020
- Part-Aware Prototype Network for Few-Shot Semantic Segmentation2020
- Prior Guided Feature Enrichment Network for Few-Shot SegmentationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2020
- SG-One: Similarity Guidance Network for One-Shot Semantic SegmentationIEEE Transactions on Cybernetics, 2020
- Re-Caption: Saliency-Enhanced Image Captioning Through Two-Phase LearningIEEE Transactions on Image Processing, 2019
- Pixel Objectness: Learning to Segment Generic Objects Automatically in Images and VideosIEEE Transactions on Pattern Analysis and Machine Intelligence, 2018
- Scalable scene understanding via saliency consensusSoft Computing, 2017
- CPMC: Automatic Object Segmentation Using Constrained Parametric Min-CutsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2011
- TextonBoost for Image Understanding: Multi-Class Object Recognition and Segmentation by Jointly Modeling Texture, Layout, and ContextInternational Journal of Computer Vision, 2007