Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Few-Shot Segmentation via Cycle-Consistent Transformer

About

Few-shot segmentation aims to train a segmentation model that can fast adapt to novel classes with few exemplars. The conventional training paradigm is to learn to make predictions on query images conditioned on the features from support images. Previous methods only utilized the semantic-level prototypes of support images as conditional information. These methods cannot utilize all pixel-wise support information for the query predictions, which is however critical for the segmentation task. In this paper, we focus on utilizing pixel-wise relationships between support and query images to facilitate the few-shot segmentation task. We design a novel Cycle-Consistent TRansformer (CyCTR) module to aggregate pixel-wise support features into query ones. CyCTR performs cross-attention between features from different images, i.e. support and query images. We observe that there may exist unexpected irrelevant pixel-level support features. Directly performing cross-attention may aggregate these features from support to query and bias the query features. Thus, we propose using a novel cycle-consistent attention mechanism to filter out possible harmful support features and encourage query features to attend to the most informative pixels from support images. Experiments on all few-shot segmentation benchmarks demonstrate that our proposed CyCTR leads to remarkable improvement compared to previous state-of-the-art methods. Specifically, on Pascal-$5^i$ and COCO-$20^i$ datasets, we achieve 67.5% and 45.6% mIoU for 5-shot segmentation, outperforming previous state-of-the-art methods by 5.6% and 7.1% respectively.

Gengwei Zhang, Guoliang Kang, Yi Yang, Yunchao Wei• 2021

Related benchmarks

TaskDatasetResultRank
Few-shot SegmentationPASCAL-5i
mIoU (Fold 0)73.5
325
Few-shot Semantic SegmentationPASCAL-5^i (test)
FB-IoU75.4
177
Few-shot SegmentationCOCO 20^i (test)
mIoU45.6
174
Semantic segmentationCOCO-20i
mIoU (Mean)45.6
132
Few-shot Semantic SegmentationCOCO-20i
mIoU45.6
115
Semantic segmentationPASCAL-5i
Mean mIoU67.5
111
Few-shot Semantic SegmentationPASCAL-5i
mIoU66.6
96
Few-shot Semantic SegmentationCOCO 5-shot 20i
mIoU45.6
85
Few-shot SegmentationPASCAL 5i (val)
mIoU (Mean)67.5
83
Few-shot Semantic SegmentationCOCO-20i (test)
mIoU (mean)45.6
79
Showing 10 of 34 rows

Other info

Follow for update