.png){kind=icon}
返回
A complementary and contrastive network for stimulus segmentation and generalization
.png){kind=icon}
PRE
AI
DOI：10.1016/j.imavis.2023.104694.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Existing convolutional neural networks (CNNs) have achieved remarkable performance in medical image segmentation tasks. However, they still fail to generalize well to unseen datasets due to the limited size and diversity of training data as well as distribution shifts. Meanwhile, CNN-based methods have inherent limi-tations in capturing global contexts and suffer semantic dilution issues in the decoder stage, which leads to sub-optimal predictions especially under low inter-class discrepancy and complex backgrounds. In this paper, we propose a novel framework named CCNet that learns complementary and contrastive features for accurate seg-mentation. Firstly, a novel complementary feature extraction module is formulated to learn global-local features by coordinating Transformer and CNN-style parallel branches. Secondly, a global context refinement module is constructed to adaptively generate a set of layer-specific global maps, so as to remedy semantic dilution. Thirdly, a mutual attentive module is designed to alleviate background confusion, in which contrastive cues are mutually captured from the foreground and background view by cascaded dual attention blocks. Moreover, we implement synthetic data augmentation to deal with training data scarcity and distribution shifts, thereby improving the out-of-distribution generalization of our model. Extensive experiments demonstrate that our CCNet achieves outstanding performance in polyp, skin lesion, and nuclei segmentation tasks, outperforming the state-of-the-arts. & COPY; 2023 Published by Elsevier B.V.
Keyword:
Complementary feature
Contrastive feature
Mutual attention
Synthetic data augmentation
Medical image segmentation
期刊
IF:
4.2
论文数:
4.1K
被引数:
6.7K
机构
引用论文
Inflexibility of mental planning: A characteristic disorder with prefrontal lobe lesions?心理计划的僵化: 前额叶病变的特征性障碍？
PREAI
分享收藏WM-DOVA maps for accurate polyp highlighting in colonoscopy: Validation vs. saliency maps from physicians在结肠镜检查中准确突出显示息肉的wm-dova地图: 医生的验证与显著性地图
PREAI
分享收藏Deep co-supervision and attention fusion strategy for automatic COVID-19 lung infection segmentation on CT images深度协同监督和注意力融合策略在CT图像新型冠状病毒肺炎肺部感染自动分割中的应用.png)
PATTERN RECOGNITION
OAAI
PATTERN RECOGNITIONIF7.6
OAAI分享收藏Satellite‐derived NDVI underestimates the advancement of alpine vegetation growth over the past three decadesEcologyOAAI
EcologyIF0
OAAI分享收藏
查看更多内容