Semantic Segmentation
文章平均质量分 74
Semantic Segmentation
mingo_敏
这个作者很懒,什么都没留下…
展开
-
深度学习论文: ISTDU-Net:Infrared Small-Target Detection U-Net及其PyTorch实现
深度学习论文: ISTDU-Net:Infrared Small-Target Detection U-Net及其PyTorch实现ISTDU-Net:Infrared Small-Target Detection U-NetPDF: https://doi.org/10.1109/LGRS.2022.3141584PyTorch代码: https://github.com/shanglianlm0525/CvPytorchPyTorch代码: https://github.com/shanglia原创 2023-09-20 17:01:13 · 1268 阅读 · 0 评论 -
深度学习论文: IncepFormer: Efficient Inception Transformer with Pyramid Pooling for Semantic Segmentation
InceptFormer由两个主要部分组成:1)金字塔Inception Transformer编码器用于捕获粗粒度和细粒度特征;2)轻量级上采样Concat解码器合并多尺度特征并进行piexl级预测。原创 2022-12-12 17:46:29 · 604 阅读 · 0 评论 -
深度学习论文: SegFormer:Simple and Efficient Design for Semantic Segmentation with Transformers及其PyTorch实现
SegFormer使用一种分层特征表示的方法,每个transformer层的输出特征尺寸逐层递减,通过这种方式捕获不同尺度的特征信息。并且舍弃了ViT中的position embedding操作,避免了测试图像与训练图像尺寸不同而导致模型性能下降的问题。在decoder部分采用简单的MLP结构,聚合transformer层不同尺度的特征,可以同时融合局部注意力和全局注意力。原创 2022-10-19 21:14:18 · 1512 阅读 · 1 评论 -
深度学习论文: SegNeXt: Rethinking Convolutional Attention Design for Semantic Segmentation及其PyTorch实现
1 采用强骨干网络作为编码器;2 多尺度信息交互;3 空域注意力;4 低计算复杂度。基于此,不同于已有Transformer方案,提出的SegNeXt对编码器模块采用传统卷积模块设计但引入了多尺度卷积注意力,对解码器模块采用了Hamberger(自注意力的一种替代方案)进一步提取全局上下文信息。提出的SegNeXt兼具性能和速度的优势;原创 2022-10-01 07:00:00 · 1780 阅读 · 0 评论 -
深度学习论文: Lightweight and Progressively-Scalable Networks for Semantic Segmentation及其PyTorch实现
LPSNet首先构建一个微型网络。然后,LPSNet通过一次扩展单个维度(卷积块的数量、通道的数量或输入分辨率)逐步将微型网络扩展到更大的网络,以满足最佳速度/准确性权衡。LPSNet在Cityscapes上实现了73.4%mIoU和413.5FPS的速度。......原创 2022-08-01 20:23:06 · 858 阅读 · 0 评论 -
深度学习论文: U-Net and its variants for Medical Image Segmentation : A short review
深度学习论文: U-Net and its variants for Medical Image Segmentation : A short reviewU-Net and its variants for Medical Image Segmentation : A short reviewPDF: https://arxiv.org/pdf/2204.08470v1.pdfPyTorch代码: https://github.com/shanglianlm0525/CvPytorchPyTorc原创 2022-04-24 10:57:18 · 2926 阅读 · 0 评论 -
深度学习论文: TopFormer: Token Pyramid Transformer for Mobile Semantic Segmentation及其PyTorch实现
深度学习论文: TopFormer: Token Pyramid Transformer for Mobile Semantic Segmentation及其PyTorch实现TopFormer: Token Pyramid Transformer for Mobile Semantic SegmentationPDF: https://arxiv.org/pdf/2204.05525.pdfPyTorch代码: https://github.com/shanglianlm0525/CvPytorch原创 2022-04-22 12:14:19 · 2206 阅读 · 0 评论 -
深度学习论文: Real-time Semantic Segmentation via Spatial-detail Guided Context Propagation
深度学习论文: Real-time Semantic Segmentation via Spatial-detail Guided Context PropagationReal-time Semantic Segmentation via Spatial-detail Guided Context PropagationPDF: https://arxiv.org/pdf/2005.11034v4.pdfPyTorch代码: https://github.com/shanglianlm0525/Cv原创 2022-04-13 09:34:25 · 3367 阅读 · 0 评论 -
深度学习论文: Rethink Dilated Convolution for Real-time Semantic Segmentation及其PyTorch实现
深度学习论文: Rethink Dilated Convolution for Real-time Semantic Segmentation及其PyTorch实现Rethink Dilated Convolution for Real-time Semantic SegmentationPDF: https://arxiv.org/pdf/2111.09957.pdfPyTorch代码: https://github.com/shanglianlm0525/CvPytorch原创 2021-12-02 16:39:03 · 3925 阅读 · 2 评论 -
深度学习论文: NAM: Normalization-based Attention Module及其PyTorch实现
深度学习论文: NAM: Normalization-based Attention Module及其PyTorch实现NAM: Normalization-based Attention ModulePDF: https://arxiv.org/pdf/2111.12419.pdfPyTorch代码: https://github.com/shanglianlm0525/CvPytorchPyTorch代码: https://github.com/shanglianlm0525/PyTorch-N原创 2021-12-01 09:23:17 · 3771 阅读 · 9 评论 -
深度学习论文: Rethinking BiSeNet For Real-time Semantic Segmentation及其PyTorch实现
深度学习论文: Rethinking BiSeNet For Real-time Semantic Segmentation及其PyTorch实现Rethinking BiSeNet For Real-time Semantic SegmentationPDF:https://arxiv.org/pdf/2104.13188.pdfPyTorch: https://github.com/shanglianlm0525/PyTorch-Networks原创 2021-05-06 08:58:02 · 615 阅读 · 0 评论 -
深度学习论文: ICNet for Real-Time Semantic Segmentation on High-Resolution Images及其PyTorch实现
ICNet:ICNet for Real-Time Semantic Segmentation on High-Resolution Images (2018)PDF: https://arxiv.org/pdf/1704.08545.pdfPyTorch: https://github.com/shanglianlm0525/PyTorch-Networks1 概述ICNet是一个基于PSPNet的实时语义分割网络,设计目的是减少PSPNet推断耗时的同时保持较高的检测精度。ICNet可以在10原创 2020-11-01 09:25:30 · 1006 阅读 · 1 评论 -
深度学习论文: LEDnet: A lightweight encoder-decoder network for real-time semantic segmentation及其PyTorch实现
深度学习论文: LEDnet: A lightweight encoder-decoder network for real-time semantic segmentation及其PyTorch实现LEDnet: A lightweight encoder-decoder network for real-time semantic segmentationPDF:https://arxiv.org/pdf/1905.02423.pdfPyTorch: https://github.com/shan原创 2020-07-01 12:52:38 · 1355 阅读 · 0 评论 -
深度学习论文: A Compact Convolutional Neural Network for Surface Defect Inspection及其PyTorch实现
A Compact Convolutional Neural Network for Surface Defect InspectionPDF:https://www.mdpi.com/1424-8220/20/7/1974/xmlPyTorch: https://github.com/shanglianlm0525/PyTorch-Networks1 LW(LightWeight) bottleneckclass LWbottleneck(nn.Module): def __init__原创 2020-07-01 12:49:44 · 1177 阅读 · 0 评论 -
深度学习论文: Rotate to Attend: Convolutional Triplet Attention Module及其PyTorch实现
Rotate to Attend: Convolutional Triplet Attention ModulePDF:https://arxiv.org/pdf/2010.03045.pdfPyTorch: https://github.com/shanglianlm0525/PyTorch-Networks1 概述2 Triplet Attention3 Experiments3-1 ImageNet3-2 COCO原创 2020-11-01 09:16:50 · 1469 阅读 · 3 评论 -
深度学习论文: ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation及其PyTorch实现
ENet: A Deep Neural Network Architecture for Real-Time Semantic SegmentationPDF:https://arxiv.org/abs/1606.02147.pdfPyTorch: https://github.com/shanglianlm0525/PyTorch-Networks2-1 ENet initial block2-2 ENet bottleneck module原创 2020-10-22 13:25:45 · 1391 阅读 · 2 评论 -
深度学习论文: ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks及其PyTorch实现
CBNet: ECA-Net: Efficient Channel Attention for Deep Convolutional Neural NetworksPDF:https://arxiv.org/pdf/1910.03151.pdfPyTorch: https://github.com/shanglianlm0525/PyTorch-Networks1 概述作者通过比对试验分别比较了降维和通道信息交流对SE Block的影响,分别得出两个结论,不降维的SE具有更好性能,通道信息交流有助于原创 2020-10-13 10:23:34 · 2491 阅读 · 0 评论 -
深度学习论文: Fast-SCNN: Fast Semantic Segmentation Network及其PyTorch实现
深度学习论文: Fast-SCNN: Fast Semantic Segmentation Network及其PyTorch实现Fast-SCNN: Fast Semantic Segmentation NetworkPDF:https://arxiv.org/pdf/1902.04502.pdfPyTorch: https://github.com/shanglianlm0525/PyTorch-Networks1 概述Fast SCNN 受 two-branch 结构和 encoder-dec原创 2020-08-02 21:32:35 · 2213 阅读 · 2 评论 -
深度学习论文: Pyramidal Convolution: Rethinking CNN for Visual Recognition及其PyTorch实现
深度学习论文: Pyramidal Convolution: Rethinking Convolutional Neural Networks for Visual Recognition及其PyTorch实现Pyramidal Convolution: Rethinking Convolutional Neural Networks for Visual RecognitionPDF:https://arxiv.org/pdf/2006.11538.pdfPyTorch: https://githu原创 2020-07-01 12:54:10 · 2332 阅读 · 1 评论 -
深度学习论文: LRNnet: a light-weighted network for real-time semantic segmentation及其PyTorch实现
深度学习论文: LRNnet: a light-weighted network with efficient reduced non-local operation for real-time semantic segmentation及其PyTorch实现LRNnet: a light-weighted network with efficient reduced non-local operation for real-time semantic segmentationPDF:https://a原创 2020-07-01 12:51:37 · 1457 阅读 · 2 评论 -
注意力机制论文:Concurrent Spatial and Channel SE in Fully Convolutional Networks及其Pytorch实现
Concurrent Spatial and Channel ‘Squeeze & Excitation’ in Fully Convolutional NetworksPDF: https://arxiv.org/pdf/1803.02579v2.pdfPyTorch代码: https://github.com/shanglianlm0525/PyTorch-NetworksPyT...原创 2020-02-26 14:06:12 · 2411 阅读 · 0 评论 -
语义分割论文:DeepLabv3+: Encoder-Decoder with Atrous Separable Convolution及其Pytorch实现
DeepLabv3+: Encoder-Decoder with Atrous Separable ConvolutionPDF: https://arxiv.org/pdf/1802.02611.pdfPyTorch代码: https://github.com/shanglianlm0525/PyTorch-Networks1 创新点为了融合多尺度信息,引入encoder-decode...原创 2020-08-28 20:37:16 · 1491 阅读 · 0 评论 -
语义分割论文:RGPNET: a real-time general purpose semantic segmentation
RGPNET: A REAL-TIME GENERAL PURPOSE SEMANTIC SEGMENTATIONPDF: https://arxiv.org/pdf/1912.01394.pdfPyTorch代码: https://github.com/shanglianlm0525/PyTorch-Networks原创 2020-02-11 10:56:24 · 775 阅读 · 0 评论 -
语义分割论文:LiteSeg: A Novel Lightweight ConvNet for Semantic Segmentation及其Pytorch实现
LiteSeg: A Novel Lightweight ConvNet for Semantic SegmentationPDF: https://arxiv.org/pdf/1912.06683.pdfPyTorch代码: https://github.com/shanglianlm0525/PyTorch-Networks原创 2020-02-10 16:47:52 · 3697 阅读 · 1 评论 -
语义分割论文:SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation (arXiv2015)
SegNet:SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation (arXiv2015)https://arxiv.org/pdf/1511.00561.pdfhttp://mi.eng.cam.ac.uk/projects/segnet/tutorial.html特点 :使用...原创 2019-04-01 13:14:06 · 794 阅读 · 0 评论 -
语义分割论文:U-Net: Convolutional Networks for Biomedical Image Segmentation (MICCAI2015)
U-Net:U-Net: Convolutional Networks for Biomedical Image Segmentation (MICCAI2015)https://arxiv.org/pdf/1505.04597.pdfhttps://lmb.informatik.uni-freiburg.de/people/ronneber/u-net/PyTorch: https://...原创 2019-04-01 13:14:32 · 1441 阅读 · 0 评论 -
语义分割论文:Learning Deconvolution Network for Semantic Segmentation (ICCV2015)
Deconvolution:Learning Deconvolution Network for Semantic Segmentation (ICCV2015)https://arxiv.org/pdf/1505.04366.pdfhttp://cvlab.postech.ac.kr/research/deconvnet/Caffe: https://github.com/Hyeonw...原创 2019-04-01 13:15:59 · 1027 阅读 · 0 评论 -
语义分割论文:Fully Convolutional Networks for Semantic Segmentation (CVPR2015)
FCN:Fully Convolutional Networks for Semantic Segmentation (CVPR2015)https://arxiv.org/pdf/1411.4038.pdfPyTorch: https://github.com/ZijunDeng/pytorch-semantic-segmentation第一篇将CNN引入image segmentat...原创 2019-04-01 13:15:43 · 718 阅读 · 0 评论 -
语义分割论文:Pyramid Scene Parsing Network (2017)
PSPNet:Pyramid Scene Parsing Network (2017)https://arxiv.org/pdf/1612.01105.pdfhttps://hszhao.github.io/projects/pspnet/特点:能够有效的得到像素级场景标注的全局上下文信息, pyramid pooling module 集合不同 levels 的信息,比 global...原创 2019-04-01 10:38:11 · 612 阅读 · 0 评论 -
语义分割论文:Context Encoding for Semantic Segmentation (CVPR2018)
EncNet:Context Encoding for Semantic Segmentation (2018)https://arxiv.org/pdf/1803.08904.pdfhttps://github.com/zhanghang1989/PyTorch-Encoding原创 2019-04-08 11:26:56 · 1486 阅读 · 0 评论 -
语义分割论文:Dilated Residual Networks (CVPR2017)
DRN:Dilated Residual Networks (CVPR2017)https://arxiv.org/pdf/1705.09914.pdfPytorch: https://github.com/fyu/drn特点:本文是作者将残差网络Deep Residual Networks (Deep Residual Learning for Image Recognition ...原创 2019-04-01 13:16:53 · 978 阅读 · 0 评论 -
语义分割论文:ShuffleSeg: Real-time Semantic Segmentation Network (arxiv2018)
ShuffleSeg:ShuffleSeg: Real-time Semantic Segmentation Network (arxiv2018)https://arxiv.org/pdf/1803.03816.pdfTensorFlow: https://github.com/MSiam/TFSegmentation特点:基于ShuffleNet (Shufflenet: An e...原创 2019-04-01 13:15:15 · 1563 阅读 · 0 评论 -
语义分割论文:ESPNet:Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation (ECCV2018)
ESPNet:ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation (ECCV2018)https://arxiv.org/pdf/1803.06815.pdfPyTorch: https://github.com/sacmehta/ESPNet原创 2019-04-01 10:38:41 · 1421 阅读 · 0 评论 -
语义分割论文:ESPNetv2: A Light-weight, Power Efficient, and General Purpose Convolutional Neural Network
ESPNetv2:ESPNetv2: A Light-weight, Power Efficient, and General Purpose Convolutional Neural Network (CVPR2019)https://arxiv.org/pdf/1811.11431.pdfPyTorch: https://github.com/sacmehta/ESPNetv2原创 2019-04-01 10:38:49 · 2166 阅读 · 0 评论 -
语义分割论文:BiSeNet: Bilateral Segmentation Network for Real-time Semantic Segmentation (ECCV2018)
BiSeNet:BiSeNet: Bilateral Segmentation Network for Real-time Semantic Segmentation (ECCV2018)https://arxiv.org/pdf/1808.00897.pdf特点:1)Spatial Path 以保留原输入图像的空间尺度,并编码丰富的空间信息。Spatial Path 包含三层,每层包含一...原创 2019-04-01 10:37:48 · 1376 阅读 · 0 评论 -
语义分割论文:Dual Attention Network for Scene Segmentation (CVPR2019)
DANDual Attention Network for Scene Segmentation (CVPR2019)https://arxiv.org/pdf/1809.02983.pdf[Pytorch] https://github.com/junfu1115/DANet特点:提出了Dual Attention Networks (DANet)在spatial和channle维度...原创 2019-04-01 10:38:29 · 2512 阅读 · 0 评论 -
语义分割论文:FastFCN:Rethinking Dilated Convolution in the Backbone for Semantic Segmentation(arxiv2019)
FastFCN:FastFCN:Rethinking Dilated Convolution in the Backbone for Semantic Segmentation (arxiv2019)https://arxiv.org/pdf/1903.11816.pdf[PyTorch]: https://github.com/wuhuikai/FastFCN特点:1)2)3)4...原创 2019-04-08 11:27:23 · 4440 阅读 · 2 评论