综述
主要参考这篇博客Semantic Segmentation using Fully Convolutional Networks over the years其中列举了从FCN网络开始的最新的语义分割网络的相关论文和实现。这篇博客同时提供了一篇综述A Review on Deep Learning Techniques Applied to Semantic Segmentation,下面是列举的实现的文中语义分割的pytorch代码实现: pytorch-semseg
Exploring semantic segmentation with deep learning 这篇文章也列举了很多语义分割网络结构。
SemanticSegmentation_DL 这是github上一个科研作者一直更新的语义分割相关论文列表,可以参考。 推荐指数 *****
下文中也有分类好的语义分割网络论文 2015-10-09-segmentation.md
A Review on Deep Learning Techniques Applied to Semantic Segmentation 该论文主要介绍了基于深度学习进行语义分割的相关方法,具有一定的参考价值,下面是翻译的文章。
综述论文翻译:A Review on Deep Learning Techniques Applied to Semantic Segmentation
Computer Vision for Autonomous Vehicles: Problems, Datasets and State-of-the-Art Kitti提供的综述文章,其中一章语义分割可以细看
该文章主要是在空中使用深度学习进行语义分割 Deep Learning for Semantic Segmentation of Aerial Imagery
SemanticSegmentation_DL 该博客整理了相关语义分割的论文和代码。
相关博客
deep learning for aerial/satellite imagery
深度学习从入门到放弃之CV-Semantic Segmentation目录 介绍了作者对部分经典语义分割框架的理解,推荐指数**
LearnSegmentation 实现了部分算法,仅仅部分参考。
反卷积(Deconvolution)、上采样(UNSampling)与上池化(UnPooling) 其中对语义分割中常用的deconvolution、unpooling和unsampling层进行了图示说明。
DeepLEGO 将语义分割各个部分分类组成,像乐高一样组成语义分割网络。
硕士博士论文
Per-Pixel Feedback for improving Semantic Segmentation
相关论文
实例分割
- instance-segmentation-pytorch Semantic Instance Segmentation with a Discriminative Loss Function
- [COCO 2018 Panoptic Segmentation Task])(http://cocodataset.org/#panoptic-2018) 全景分割比赛。
全监督
- Gated Feedback Refinement Network for Dense Image Labeling
- RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation
- ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation
- LinkNet: Exploiting Encoder Representations for Efficient Semantic Segmentation
- Learning Deconvolution Network for Semantic Segmentation 论文
- V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation vnet.pytorch实现
- U-Net: Convolutional Networks for Biomedical Image Segmentation
- DeepUNet: A Deep Fully Convolutional Network for Pixel-level Sea-Land Segmentation
- Segmentation from Natural Language Expressions
- Feedforward semantic segmentation with zoom-out features
- Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs pytorch代码 deeplab-pytorch代码
- Rethinking Atrous Convolution for Semantic Image Segmentation 论文
- Conditional Random Fields as Recurrent Neural Networks 论文
- Efficient piecewise training of deep structured models for semantic segmentation 论文
- ParseNet: Looking Wider to See Better 论文
- Semantic Image Segmentation via Deep Parsing Network 论文
- Multi-Scale Context Aggregation by Dilated Convolutions 论文 代码 当前最好的语义分割网络
- Predicting Scene Parsing and Motion Dynamics in the Future 论文 本文进行了语义分割并同时进行运动物体的预测
- Material Recognition in the Wild with the Materials in Context Database 论文
- Predicting Depth, Surface Normals and Semantic Labels with a Common Multi-Scale Convolutional Architecture 论文
- ReSeg: A Recurrent Neural Network-based Model for Semantic Segmentation 论文 代码 试验包括CamVid
- LSTM-CF: Unifying Context Modeling and Fusion with LSTMs for RGB-D Scene Labeling 论文
- Squeeze-SegNet: A new fast Deep Convolutional Neural Network for Semantic Segmentation 论文 较新的论文
- Efficient ConvNet for Real-time Semantic Segmentation
- ERFNet- Efficient Residual Factorized ConvNet for Real-time Semantic Segmentation
- Semantic Segmentation via Structured Patch Prediction, Context CRF and Guidance CRF 代码
- Semantic Instance Segmentation with a Discriminative Loss Function 论文
- Fast Scene Understanding for Autonomous Driving 论文 代码 无人驾驶领域上的快速场景理解,同时实现语义分割,实例分割和单目深度估计,可以借鉴同时实现语义分割,视频预测
- Understanding Convolution for Semantic Segmentation 论文 代码 图森科技在语义分割系统上的改进,提出了DUC和HDC。
- In-Place Activated BatchNorm for Memory-Optimized Training of DNNs 论文 本文提出了新的BatchNorm方法,加速网络的训练。 代码
- Instance Embedding Transfer to Unsupervised Video Object Segmentation 论文 使用视频进行非监督物体分割。
- Panoptic Segmentation 论文 凯明提出的新论文,重点关注。
- Stacked Deconvolutional Network for Semantic Segmentation 论文 使用类似DenseNet思路栈式构造装置卷积网络实现,试验包括CamVid
- S4Net: Single Stage Salient-Instance Segmentation 论文 代码
- Semantic Segmentation via Highly Fused Convolutional Network with Multiple Soft Cost Functions
- Convolutional CRFs for Semantic Segmentation 论文 将CRF结合到卷积中。思路非常简单,可以阅读文章参考ConvCRF代码。
- RTSeg: Real-time Semantic Segmentation Comparative Study 实时语义分割比较研究,值得阅读参考,代码参考TFSegmentation。
- ShuffleSeg: Real-time Semantic Segmentation Network 使用shuffle网络结构加速语义分割。
- AdaptSegNet Learning to Adapt Structured Output Space for Semantic Segmentation, CVPR 2018 (spotlight)从一个domain中训练自适应到另一个domain中的语义分割网络训练方法。可参考Visual Domain Adaptation Challenge
- The Lovász-Softmax loss: A tractable surrogate for the optimization of the intersection-over-union measure in neural networks,一种新的loss训练语义分割jaccardSegment
- Joint scene classification and semantic segmentation with FuseNetFuseNet_PyTorch代码实现
- ESPNet
- PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation 3D分类和分割,以及对应的代码pointnet
- Context Encoding for Semantic Segmentation
- PixelLink: Detecting Scene Text via Instance Segmentation
- Y-Net: Joint Segmentation and Classification for Diagnosis of Breast Biopsy Images 类似于U-Net在癌症语义分割中的应用。
- Learning to Adapt Structured Output Space for Semantic Segmentation AdaptSegNet代码
- DenseASPP for Semantic Segmentation in Street Scenes DenseASPP代码
- Learning to Segment Every Thing seg_every_thing论文代码
- Learning from Synthetic Data: Addressing Domain Shift for Semantic Segmentation 这被称为域适应语义分割,也就是从不同域分布的数据集中训练的模型转换到新数据中的模型,代码LSD-seg
- Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
- Concurrent Spatial and Channel Squeeze & Excitation in Fully Convolutional Networks
- Vortex Pooling: Improving Context Representation in Semantic Segmentation
弱监督
- Semi and Weakly Supervised Semantic Segmentation Using Generative Adversarial Network 论文
- Decoupled Deep Neural Network for Semi-supervised Semantic Segmentation
- Semi-Supervised Learning with Ladder Networks
- Feedback Neural Network for Weakly Supervised Geo-Semantic Segmentation 论文
- Weakly Supervised Semantic Segmentation with Convolutional Networks 论文
- BoxSup: Exploiting Bounding Boxes to Supervise Convolutional Networks for Semantic Segmentation 论文 使用bounding box进行语义分割
- ScribbleSup: Scribble-Supervised Convolutional Networks for Semantic Segmentation 论文
- One-Shot Learning for Semantic Segmentation 论文
- Semantic Segmentation from Limited Training Data 论文
- Weakly- and Semi-Supervised Learning of a DCNN for Semantic Image Segmentation 论文
- Weakly Supervised Semantic Segmentation using Web-Crawled Videos 论文
- Fully Convolutional Multi-Class Multiple Instance Learning 论文
- Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks 论文
- Seed, Expand and Constrain: Three Principles for Weakly-Supervised Image Segmentation 论文 代码 模型数据
- Constrained Convolutional Neural Networks for Weakly Supervised Segmentation 论文 代码
- Semantic segmentation using adversarial networks 论文 代码 Semantic-Segmentation-using-Adversarial-Networks
- Adversarial Deep Structural Networks for Mammographic Mass Segmentation 论文
- W-Net: A Deep Model for Fully Unsupervised Image Segmentation 论文
- sscnet Semantic Scene Completion from a Single Depth Image一张深度图中获取语义场景填充。
- TernausNetV2 TernausNetV2: Fully Convolutional Network for Instance Segmentation
- Learning Pixel-Level Semantic Affinity With Image-Level Supervision for Weakly Supervised Semantic Segmentation
- Learning From Synthetic Data: Addressing Domain Shift for Semantic Segmentation
- DenseASPP for Semantic Segmentation in Street Scenes
- Context Contrasted Feature and Gated Multi-Scale Aggregation for Scene Segmentation
弱监督相关博客
本文介绍了4篇弱监督学习的论文 - CNN在基于弱监督学习的图像分割中的应用 - WeakSupervisedSegmentationList 弱监督分割论文列表。
常用框架
pytorch
- pytorch-semantic-segmentation 用于收集常用的语义分割的框架,比如FCN、U-Net、SegNet、PSPNet、GCN、DUC和HDC pytorch-semantic-segmentation
开源实现
Tensorflow-Segmentation 使用tensorflow实现了SegNet等编码器解码器分割网络。
PixelAnnotationTool 用来像素标注的工具,和LabelMe在图像分类和检测中的作用类似。
相关开源项目
Adversarial-Semisupervised-Semantic-Segmentation 基于对抗生成网络的语义分割网络
pixel_level_land_classification 用于对空中图像land语义分割的仓库。