Concurrent Spatial and Channel ‘Squeeze &amp; Excitation’ in Fully Convolutional Networks

Roy, Abhijit Guha; Navab, Nassir; Wachinger, Christian

doi:10.1007/978-3-030-00928-1_48

Cited by 674 publications

(391 citation statements)

References 8 publications

Supporting

Mentioning

389

Contrasting

Unclassified

Order By: Relevance

“…The SE module first squeezes the feature map by global average pooling and then passes the squeezed feature to the gating module to get the representation of channel-wise dependencies, which is used to re-calibrate the feature map to emphasize on useful channels. The work in [10] refers to the SE module in [9] as Spatial Squeeze and Channel Excitation (cSE) and proposes a different version called Channel Squeeze and Spatial Excitation (sSE). The sSE module squeezes the feature map along channels to preserve more spatial information, thus is more suitable for image segmentation task.…”

Section: Squeeze and Excitationmentioning

confidence: 99%

Mixed-Supervised Dual-Network for Medical Image Segmentation

Wang

Ben‐Shlomo

et al. 2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Deep learning based medical image segmentation models usually require large datasets with high-quality dense segmentations to train, which are very time-consuming and expensive to prepare. One way to tackle this challenge is using the mixed-supervised learning framework, in which only a part of data is densely annotated with segmentation label and the rest is weakly labeled with bounding boxes. The model is trained jointly in a multi-task learning setting. In this paper, we propose Mixed-Supervised Dual-Network (MSDN), a novel architecture which consists of two separate networks for the detection and segmentation tasks respectively, and a series of connection modules between the layers of the two networks. These connection modules are used to transfer useful information from the auxiliary detection task to help the segmentation task. We propose to use a recent technique called Squeeze and Excitation in the connection module to boost the transfer. We conduct experiments on two medical image segmentation datasets. The proposed MSDN model outperforms multiple baselines.

show abstract

Section: Squeeze and Excitationmentioning

confidence: 99%

Mixed-Supervised Dual-Network for Medical Image Segmentation

Wang

Ben‐Shlomo

et al. 2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…Attention is widely used for re-weighting features with high-level information in deep networks. Roy et al applied a concurrent attention module to semantic segmentation [14], where features were squeezed along spatial and channel axes and then applied to the original feature maps to provide spatial and contextual information. EncNet [15] introduces a context encoding module at the end of the network, to encode global contextual information and re-weight the extracted features for discriminative representations.…”

Section: Related Workmentioning

confidence: 99%

“…The spatial attention block takes the fused features and the output of the spatial branch as input, it helps refine the pixel localizations and object boundaries. Similar to the attention module in [14], the features from the spatial branch go through a 3×3 convolution with batch normalization and Sigmoid non-linearity, and then multiplied by the fused features. Fig.…”

Section: Feature Cross Attention Modulementioning

confidence: 99%

Cross Attention Network for Semantic Segmentation

Liu

Yin

2019

2019 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

In this paper, we address the semantic segmentation task with a deep network that combines contextual features and spatial information. The proposed Cross Attention Network is composed of two branches and a Feature Cross Attention (FCA) module. Specifically, a shallow branch is used to preserve low-level spatial information and a deep branch is employed to extract high-level contextual features. Then the FCA module is introduced to combine these two branches. Different from most existing attention mechanisms, the FCA module obtains spatial attention map and channel attention map from two branches separately, and then fuses them. The contextual features are used to provide global contextual guidance in fused feature maps, and spatial features are used to refine localizations. The proposed network outperforms other realtime methods with improved speed on the Cityscapes and CamVid datasets with lightweight backbones, and achieves state-of-the-art performance with a deep backbone.

show abstract

“…For comparison, the U-Net [4], U-Net++ [11], and SE U-Net [12] were trained and finetuned using same loss function and the proposed weakly supervised training strategy. Table 2 lists the segmentation accuracy measured by mean Dice score of all the testing groups.…”

Section: Glomerular Segmentationmentioning

confidence: 99%

Towards Staining Independent Segmentation of Glomerulus from Histopathological Images of Kidney

Liu

Wang

et al. 2019

Preprint

View full text Add to dashboard Cite

This paper introduces a detection-based framework to segment glomeruli from digital scanning image of light microscopic slide of renal biopsy specimens. The proposed method aims to better use the precise localization ability of Faster R-CNN and powerful segmentation ability of U-Net. We use a detector to localize the glomeruli from whole slide image to make the segmentation only focus on the most relevant area of the image. We explored the effectiveness of the network depth on its localization and segmentation ability in glomerular classification, and then propose to use the classification network with enhanced ability of localization and segmentation to construct and initialize a segmentation network. We also propose a weakly supervised training strategy to train the segmentation network by taking advantage of the unique morphology of the glomerulus. Both strong initialization and weakly supervised training are used to resolve the problem of insufficient and inaccurate data annotations and enhance the adaptability of the segmentation network. Experimental results demonstrate that the proposed framework is effective and robust.

show abstract

Concurrent Spatial and Channel ‘Squeeze & Excitation’ in Fully Convolutional Networks

Cited by 674 publications

References 8 publications

Mixed-Supervised Dual-Network for Medical Image Segmentation

Mixed-Supervised Dual-Network for Medical Image Segmentation

Cross Attention Network for Semantic Segmentation

Towards Staining Independent Segmentation of Glomerulus from Histopathological Images of Kidney

Contact Info

Product

Resources

About