26,568 Hits in 5.5 sec

Expectation-Maximization Attention Networks for Semantic Segmentation [article]

Xia Li, Zhisheng Zhong, Jianlong Wu, Yibo Yang, Zhouchen Lin, Hong Liu
2019 arXiv   pre-print
The proposed Expectation-Maximization Attention (EMA) module is robust to the variance of input and is also friendly in memory and computation.  ...  In this paper, we formulate the attention mechanism into an expectation-maximization manner and iteratively estimate a much more compact set of bases upon which the attention maps are computed.  ...  Expectation-Maximization Attention In view of the high computational complexity of the attention mechanism and limitations of the Non-local module, we first propose the expectation-maximization attention  ... 
arXiv:1907.13426v2 fatcat:fqzzfqxcmnd55mbupmfsspdzhq

Edge Guided Context Aggregation Network for Semantic Segmentation of Remote Sensing Imagery

Zhiqiang Liu, Jiaojiao Li, Rui Song, Chaoxiong Wu, Wei Liu, Zan Li, Yunsong Li
2022 Remote Sensing  
modules: edge extraction module (EEM), dual expectation maximization attention module (DEMA), and edge guided module (EGM).  ...  In this paper, a novel edge guided context aggregation network (EGCAN) is proposed for the semantic segmentation of RSI. The Unet is employed as backbone.  ...  Dual Expectation Maximization Attention Module (DEMA) The ddge guided context aggregation branch introduces the expectation maximization algorithm into the self-attention mechanism, which runs the attention  ... 
doi:10.3390/rs14061353 fatcat:6m3vdsrvkreulohrzs63g34gie

Bottom-Up Top-Down Cues for Weakly-Supervised Semantic Segmentation [chapter]

Qibin Hou, Daniela Massiceti, Puneet Kumar Dokania, Yunchao Wei, Ming-Ming Cheng, Philip H. S. Torr
2018 Lecture Notes in Computer Science  
Our method uses deep convolutional neural networks (CNNs) and adopts an Expectation-Maximization (EM) based approach.  ...  We consider the task of learning a classifier for semantic segmentation using weak supervision in the form of image labels which specify the object classes present in the image.  ...  Motivated by this, we use an Expectation-Maximization (EM) [10, 25] based approach for weaklysupervised semantic segmentation using only image labels.  ... 
doi:10.1007/978-3-319-78199-0_18 fatcat:oexpqazvizfu3gq3apvyftfq4m

Bottom-Up Top-Down Cues for Weakly-Supervised Semantic Segmentation [article]

Qinbin Hou, Puneet Kumar Dokania, Daniela Massiceti, Yunchao Wei, Ming-Ming Cheng, Philip Torr
2017 arXiv   pre-print
Our method uses deep convolutional neural networks (CNNs) and adopts an Expectation-Maximization (EM) based approach.  ...  We consider the task of learning a classifier for semantic segmentation using weak supervision in the form of image labels which specify the object classes present in the image.  ...  Motivated by the above-mentioned facts and the work by [29] , we use an Expectation-Maximization (EM) [10, 27] based approach for weakly-supervised semantic segmentation with only image labels.  ... 
arXiv:1612.02101v3 fatcat:uvccgkkew5eqnkm4fx3e6kt5zi

Exploring Attention-Aware Network Resource Allocation for Customized Metaverse Services [article]

Hongyang Du, Jiacheng Wang, Dusit Niyato, Jiawen Kang, Zehui Xiong, Xuemin Shen, Dong In Kim
2022 arXiv   pre-print
With the help of UOAL, we propose an attention-aware network resource allocation algorithm that has two steps, i.e., attention prediction and QoE maximization.  ...  By using the predicted user-object-attention values, network resources such as the rendering capacity of edge devices can be allocated optimally to maximize the QoE.  ...  INTRODUCTION The concept of "Metaverse" embodies our good expectations for the future of the Internet [1] .  ... 
arXiv:2208.00369v1 fatcat:ylg2j4jx6rck7hgwmiw4gfbk4e

Gated Dense Convolutional Neural Networks for Unbalanced Representations in STEM Tomography

Arda Genc, Libor Kovarik, Hamish L. Fraser
2022 Microscopy and Microanalysis  
expectation maximization (MLEM) reconstructions.  ...  for Pt nanoparticle segmentation with attention units.  ... 
doi:10.1017/s1431927622011667 fatcat:63dzafghqja5ditkj7pbfqqks4

A Lightweight Semantic Segmentation Model of Wucai Seedlings Based on Attention Mechanism

Wen Li, Chao Liu, Minhui Chen, Dequan Zhu, Xia Chen, Juan Liao
2022 Photonics  
In addition, the expectation "maximizationexpectation" maximization attention module is added to enhance the attention of the model to the segmentation object.  ...  To segment wucai seedlings accurately in a natural environment, this paper presents a lightweight segmentation model of wucai seedlings, where U-Net is used as the backbone network.  ...  Expectation Maximization Attention Module Attention mechanism, widely used for segmentation tasks, is essential to make the segmentation model pay different attention to different parts of input images  ... 
doi:10.3390/photonics9060393 fatcat:vx6i33q3unaz5kplqzxpou6yjq

MVIP 2020 Table of Contents

2020 2020 International Conference on Machine Vision and Image Processing (MVIP)  
An Efficient Approach for Using Expectation Maximization Algorithm in Capsule Networks 22. Region Proposal Generation: A Hierarchical Merging Similarity-Based Algorithm 23.  ...  A DNN-based Image Retrieval Approach for Detection of Defective Area in Carbon Fiber Reinforced Polymers through LDV Data 40. Class Attention Map Distillation for Efficient Semantic Segmentation 41.  ... 
doi:10.1109/mvip49855.2020.9116904 fatcat:6v7rolxpkfh6jb2fg2bhd4ssuq

Improving Gradient Flow with Unrolled Highway Expectation Maximization [article]

Chonghyuk Song, Eunseok Kim, Inwook Shim
2020 arXiv   pre-print
In particular, many works have employed the expectation maximization (EM) algorithm in the form of an unrolled layer-wise structure that is jointly trained with a backbone neural network.  ...  To address this issue, we propose Highway Expectation Maximization Networks (HEMNet), which is comprised of unrolled iterations of the generalized EM (GEM) algorithm based on the Newton-Rahpson method.  ...  EMANet ) designs an EM-based attention module that boosts the performance of a backbone network on semantic segmentation.  ... 
arXiv:2012.04926v1 fatcat:yauhsyhhrze5hnl5jkvyqsw6qy

Learning Effectively from Noisy Supervision for Weakly Supervised Semantic Segmentation

Wenbin Xie, Qiaoqiao Wei, Zheng Li, Hui Zhang
2020 British Machine Vision Conference  
In this paper, we propose a novel framework for weakly supervised semantic segmentation (WSSS) using bounding boxes to alleviate the need for pixel-wise annotations.  ...  Therefore, we present a constrained foreground segmentation network (CFS) to generate high-quality dense annotations from noisy proposals.  ...  Wen Zheng and Huijuan Huang for their suggestion and help.  ... 
dblp:conf/bmvc/XieWLZ20 fatcat:umvl7csjfvgvdbw3w5viohz4xi

Bi-SANet—Bilateral Network with Scale Attention for Retinal Vessel Segmentation

Yun Jiang, Huixia Yao, Zeqi Ma, Jingyao Zhang
2021 Symmetry  
The segmentation of retinal vessels is critical for the diagnosis of some fundus diseases.  ...  In this paper, we proposed a two-branch network based on multi-scale attention to alleviate the above problem.  ...  Institutional Review Board Statement: Ethical review and approval are not applicable for this paper. Informed Consent Statement: An informed consent statement is not applicable.  ... 
doi:10.3390/sym13101820 fatcat:neo5kl5yijc4hobyrzwgq5wrkq

Unsupervised Representation for Semantic Segmentation by Implicit Cycle-Attention Contrastive Learning

Bo Pang, Yizhuo Li, Yifan Zhang, Gao Peng, Jiajun Tang, Kaiwen Zha, Jiefeng Li, Cewu Lu
We study the unsupervised representation learning for the semantic segmentation task.  ...  Different from previous works that aim at providing unsupervised pre-trained backbones for segmentation models which need further supervised fine-tune, here, we focus on providing representation that is  ...  With the results of the above analysis, we propose the cycle-attention contrastive learning (CACL) for semantic segmentation.  ... 
doi:10.1609/aaai.v36i2.20100 fatcat:opttcfd54fbpdbszccbndgebme

Abdominal multi-organ segmentation with organ-attention networks and statistical fusion

Yan Wang, Yuyin Zhou, Wei Shen, Seyoun Park, Elliot K. Fishman, Alan L. Yuille
2019 Medical Image Analysis  
To address these challenges, we introduce a novel framework for multi-organ segmentation of abdominal regions by using organ-attention networks with reverse connections (OAN-RCs) which are applied to 2D  ...  Accurate and robust segmentation of abdominal organs on CT is essential for many clinical applications such as computer-aided diagnosis and computer-aided surgery.  ...  The basic fusion framework uses expectation-maximization (EM) similar to [2, 36] .  ... 
doi:10.1016/ pmid:31035060 fatcat:vvaak3uwdzddxcm2sg7k2qutra

Associating Inter-image Salient Instances for Weakly Supervised Semantic Segmentation [chapter]

Ruochen Fan, Qibin Hou, Ming-Ming Cheng, Gang Yu, Ralph R. Martin, Shi-Min Hu
2018 Lecture Notes in Computer Science  
The proposed framework is general, and any state-of-the-art fully-supervised network structure can be incorporated to learn the segmentation network.  ...  When working with DeepLab for semantic segmentation, our method outperforms state-of-the-art weakly supervised alternatives by a large margin, achieving 65.6% mIoU on the PASCAL VOC 2012 dataset.  ...  China (Project Number 61521002, 61620106008, 61572264) and the Joint NSFC-ISF Research Program (project number 61561146393), the national youth talent support program, Tianjin Natural Science Foundation for  ... 
doi:10.1007/978-3-030-01240-3_23 fatcat:cl46tczynff2tdx2myeprmjvhq

SPANet: Successive Pooling Attention Network for Semantic Segmentation of Remote Sensing Images

Le Sun, Shiwei Cheng, Yuhui Zheng, Zebin Wu, Jianwei Zhang
2022 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
To alleviate this problem, a successive pooling attention network (SPANet) was proposed.  ...  In the convolutional neural network, the precise segmentation of small-scale objects and object boundaries in remote sensing images is a great challenge.  ...  For instance, in dual expectation-maximization attention network [22] , the spatial expectation-maximization attention model simulated the interdependence of spatial features to obtain rich contextual  ... 
doi:10.1109/jstars.2022.3175191 fatcat:m7fuxfjsfbeq3by5yy5krc25m4
« Previous Showing results 1 — 15 out of 26,568 results