语义分割深度学习方法集锦

转载：https://github.com/handong1587/handong1587.github.io/edit/master/_posts/deep_learning/2015-10-09-segmentation.md

Papers

Deep Joint Task Learning for Generic Object Extraction

• intro: NIPS 2014
• homepage: http://vision.sysu.edu.cn/projects/deep-joint-task-learning/
• paper: http://ss.sysu.edu.cn/~ll/files/NIPS2014_JointTask.pdf
• github: https://github.com/xiaolonw/nips14_loc_seg_testonly
• dataset: http://objectextraction.github.io/

Highly Efficient Forward and Backward Propagation of Convolutional Neural Networks for Pixelwise Classification

• arxiv: https://arxiv.org/abs/1412.4526
• code(Caffe): https://dl.dropboxusercontent.com/u/6448899/caffe.zip
• author page: http://www.ee.cuhk.edu.hk/~hsli/

Segmentation from Natural Language Expressions

• intro: ECCV 2016
• project page: http://ronghanghu.com/text_objseg/
• arxiv: http://arxiv.org/abs/1603.06180
• github(TensorFlow): https://github.com/ronghanghu/text_objseg
• gtihub(Caffe): https://github.com/Seth-Park/text_objseg_caffe

Semantic Object Parsing with Graph LSTM

• arxiv: http://arxiv.org/abs/1603.07063

Fine Hand Segmentation using Convolutional Neural Networks

• arxiv: http://arxiv.org/abs/1608.07454

Feedback Neural Network for Weakly Supervised Geo-Semantic Segmentation

• intro: Facebook Connectivity Lab & Facebook Core Data Science & University of Illinois
• arxiv: https://arxiv.org/abs/1612.02766

FusionNet: A deep fully residual convolutional neural network for image segmentation in connectomics

• arxiv: https://arxiv.org/abs/1612.05360

A deep learning model integrating FCNNs and CRFs for brain tumor segmentation

• arxiv: https://arxiv.org/abs/1702.04528

Texture segmentation with Fully Convolutional Networks

• intro: Dublin City University
• arxiv: https://arxiv.org/abs/1703.05230

Fast LIDAR-based Road Detection Using Convolutional Neural Networks

https://arxiv.org/abs/1703.03613

Deep Value Networks Learn to Evaluate and Iteratively Refine Structured Outputs

• arxiv: https://arxiv.org/abs/1703.04363
• demo: https://gyglim.github.io/deep-value-net/

Annotating Object Instances with a Polygon-RNN

• intro: CVPR 2017. CVPR Best Paper Honorable Mention Award. University of Toronto
• project page: http://www.cs.toronto.edu/polyrnn/
• arxiv: https://arxiv.org/abs/1704.05548

Semantic Segmentation via Structured Patch Prediction, Context CRF and Guidance CRF

• intro: CVPR 2017
• paper: http://openaccess.thecvf.com/content_cvpr_2017/papers/Shen_Semantic_Segmentation_via_CVPR_2017_paper.pdf
• github(Caffe): https://github.com//FalongShen/SegModel

Nighttime sky/cloud image segmentation

• intro: ICIP 2017
• arxiv: https://arxiv.org/abs/1705.10583

Distantly Supervised Road Segmentation

• intro: ICCV workshop CVRSUAD2017. Indiana University & Preferred Networks
• arxiv: https://arxiv.org/abs/1708.06118

Superpixel clustering with deep features for unsupervised road segmentation

• intro: Preferred Networks, Inc & Indiana University
• arxiv: https://arxiv.org/abs/1711.05998

Learning to Segment Human by Watching YouTube

• intro: TPAMI 2017
• arxiv: https://arxiv.org/abs/1710.01457

W-Net: A Deep Model for Fully Unsupervised Image Segmentation

https://arxiv.org/abs/1711.08506

End-to-end detection-segmentation network with ROI convolution

• intro: ISBI 2018
• arxiv: https://arxiv.org/abs/1801.02722

U-Net

U-Net: Convolutional Networks for Biomedical Image Segmentation

• intro: conditionally accepted at MICCAI 2015
• project page: http://lmb.informatik.uni-freiburg.de/people/ronneber/u-net/
• arxiv: http://arxiv.org/abs/1505.04597
• code+data: http://lmb.informatik.uni-freiburg.de/people/ronneber/u-net/u-net-release-2015-10-02.tar.gz
• github: https://github.com/orobix/retina-unet
• github: https://github.com/jakeret/tf_unet
• notes: http://zongwei.leanote.com/post/Pa

DeepUNet: A Deep Fully Convolutional Network for Pixel-level Sea-Land Segmentation

https://arxiv.org/abs/1709.00201

TernausNet: U-Net with VGG11 Encoder Pre-Trained on ImageNet for Image Segmentation

• intro: Lyft Inc. & MIT
• intro: part of the winning solution (1st out of 735) in the Kaggle: Carvana Image Masking Challenge
• arxiv: https://arxiv.org/abs/1801.05746
• github: https://github.com/ternaus/TernausNet

Foreground Object Segmentation

Pixel Objectness

• project page: http://vision.cs.utexas.edu/projects/pixelobjectness/
• arxiv: https://arxiv.org/abs/1701.05349
• github: https://github.com/suyogduttjain/pixelobjectness

A Deep Convolutional Neural Network for Background Subtraction

• arxiv: https://arxiv.org/abs/1702.01731

Semantic Segmentation

Fully Convolutional Networks for Semantic Segmentation

• intro: CVPR 2015, PAMI 2016
• keywords: deconvolutional layer, crop layer
• arxiv: http://arxiv.org/abs/1411.4038
• arxiv(PAMI 2016): http://arxiv.org/abs/1605.06211
• slides: https://docs.google.com/presentation/d/1VeWFMpZ8XN7OC3URZP4WdXvOGYckoFWGVN7hApoXVnc
• slides: http://tutorial.caffe.berkeleyvision.org/caffe-cvpr15-pixels.pdf
• talk: http://techtalks.tv/talks/fully-convolutional-networks-for-semantic-segmentation/61606/
• github(official): https://github.com/shelhamer/fcn.berkeleyvision.org
• github: https://github.com/BVLC/caffe/wiki/Model-Zoo#fcn
• github: https://github.com/MarvinTeichmann/tensorflow-fcn
• github(Chainer): https://github.com/wkentaro/fcn
• github(PyTorch): https://github.com/wkentaro/pytorch-fcn
• github(Tensorflow): https://github.com/shekkizh/FCN.tensorflow
• notes: http://zhangliliang.com/2014/11/28/paper-note-fcn-segment/

From Image-level to Pixel-level Labeling with Convolutional Networks

• intro: CVPR 2015
• intro: “Weakly Supervised Semantic Segmentation with Convolutional Networks”
• intro: performs semantic segmentation based only on image-level annotations in a multiple instance learning framework
• arxiv: http://arxiv.org/abs/1411.6228
• paper: http://ronan.collobert.com/pub/matos/2015_semisupsemseg_cvpr.pdf

Feedforward semantic segmentation with zoom-out features

• intro: CVPR 2015. Toyota Technological Institute at Chicago
• paper: http://www.cv-foundation.org/openaccess/content_cvpr_2015/papers/Mostajabi_Feedforward_Semantic_Segmentation_2015_CVPR_paper.pdf
• bitbuckt: https://bitbucket.org/m_mostajabi/zoom-out-release
• video: https://www.youtube.com/watch?v=HvgvX1LXQa8

DeepLab

Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs

• intro: ICLR 2015. DeepLab
• arxiv: http://arxiv.org/abs/1412.7062
• bitbucket: https://bitbucket.org/deeplab/deeplab-public/
• github: https://github.com/TheLegendAli/DeepLab-Context

Weakly- and Semi-Supervised Learning of a DCNN for Semantic Image Segmentation

• intro: DeepLab
• arxiv: http://arxiv.org/abs/1502.02734
• bitbucket: https://bitbucket.org/deeplab/deeplab-public/
• github: https://github.com/TheLegendAli/DeepLab-Context

DeepLab v2

DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs

• intro: TPAMI
• intro: 79.7% mIOU in the test set, PASCAL VOC-2012 semantic image segmentation task
• intro: Updated version of our previous ICLR 2015 paper
• project page: http://liangchiehchen.com/projects/DeepLab.html
• arxiv: https://arxiv.org/abs/1606.00915
• bitbucket: https://bitbucket.org/aquariusjay/deeplab-public-ver2
• github: https://github.com/DrSleep/tensorflow-deeplab-resnet
• github: https://github.com/isht7/pytorch-deeplab-resnet

DeepLabv2 (ResNet-101)

http://liangchiehchen.com/projects/DeepLabv2_resnet.html

DeepLab v3

Rethinking Atrous Convolution for Semantic Image Segmentation

• intro: Google. DeepLabv3
• arxiv: https://arxiv.org/abs/1706.05587

CRF-RNN

Conditional Random Fields as Recurrent Neural Networks

• intro: ICCV 2015. Oxford / Stanford / Baidu
• project page: http://www.robots.ox.ac.uk/~szheng/CRFasRNN.html
• arxiv: http://arxiv.org/abs/1502.03240
• github: https://github.com/torrvision/crfasrnn
• demo: http://www.robots.ox.ac.uk/~szheng/crfasrnndemo
• github: https://github.com/martinkersner/train-CRF-RNN

BoxSup

BoxSup: Exploiting Bounding Boxes to Supervise Convolutional Networks for Semantic Segmentation

• arxiv: http://arxiv.org/abs/1503.01640

Efficient piecewise training of deep structured models for semantic segmentation

• intro: CVPR 2016
• arxiv: http://arxiv.org/abs/1504.01013

DeconvNet

Learning Deconvolution Network for Semantic Segmentation

• intro: ICCV 2015. DeconvNet
• intro: two-stage training: train the network with easy examples first and
  fine-tune the trained network with more challenging examples later
• project page: http://cvlab.postech.ac.kr/research/deconvnet/
• arxiv: http://arxiv.org/abs/1505.04366
• slides: http://web.cs.hacettepe.edu.tr/~aykut/classes/spring2016/bil722/slides/w06-deconvnet.pdf
• gitxiv: http://gitxiv.com/posts/9tpJKNTYksN5eWcHz/learning-deconvolution-network-for-semantic-segmentation
• github: https://github.com/HyeonwooNoh/DeconvNet
• github: https://github.com/HyeonwooNoh/caffe

SegNet

SegNet: A Deep Convolutional Encoder-Decoder Architecture for Robust Semantic Pixel-Wise Labelling

• arxiv: http://arxiv.org/abs/1505.07293
• github: https://github.com/alexgkendall/caffe-segnet
• github: https://github.com/pfnet-research/chainer-segnet

SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation

• homepage: http://mi.eng.cam.ac.uk/projects/segnet/
• arxiv: http://arxiv.org/abs/1511.00561
• github: https://github.com/alexgkendall/caffe-segnet
• tutorial: http://mi.eng.cam.ac.uk/projects/segnet/tutorial.html

SegNet: Pixel-Wise Semantic Labelling Using a Deep Networks

• youtube: https://www.youtube.com/watch?v=xfNYAly1iXo
• mirror: http://pan.baidu.com/s/1gdUzDlD

Getting Started with SegNet

• blog: http://mi.eng.cam.ac.uk/projects/segnet/tutorial.html
• github: https://github.com/alexgkendall/SegNet-Tutorial

ParseNet

ParseNet: Looking Wider to See Better

• intro:ICLR 2016
• arxiv: http://arxiv.org/abs/1506.04579
• github: https://github.com/weiliu89/caffe/tree/fcn
• caffe model zoo: https://github.com/BVLC/caffe/wiki/Model-Zoo#parsenet-looking-wider-to-see-better

DecoupledNet

Decoupled Deep Neural Network for Semi-supervised Semantic Segmentation

• intro: ICLR 2016
• project(paper+code): http://cvlab.postech.ac.kr/research/decouplednet/
• arxiv: http://arxiv.org/abs/1506.04924
• github: https://github.com/HyeonwooNoh/DecoupledNet

Semantic Image Segmentation via Deep Parsing Network

• intro: ICCV 2015. CUHK
• keywords: Deep Parsing Network (DPN), Markov Random Field (MRF)
• homepage: http://personal.ie.cuhk.edu.hk/~lz013/projects/DPN.html
• arxiv.org: http://arxiv.org/abs/1509.02634
• paper: http://www.cv-foundation.org/openaccess/content_iccv_2015/papers/Liu_Semantic_Image_Segmentation_ICCV_2015_paper.pdf
• slides: http://personal.ie.cuhk.edu.hk/~pluo/pdf/presentation_dpn.pdf

Multi-Scale Context Aggregation by Dilated Convolutions

• intro: ICLR 2016.
• intro: Dilated Convolution for Semantic Image Segmentation
• homepage: http://vladlen.info/publications/multi-scale-context-aggregation-by-dilated-convolutions/
• arxiv: http://arxiv.org/abs/1511.07122
• github: https://github.com/fyu/dilation
• github: https://github.com/nicolov/segmentation_keras
• notes: http://www.inference.vc/dilated-convolutions-and-kronecker-factorisation/

Instance-aware Semantic Segmentation via Multi-task Network Cascades

• intro: CVPR 2016 oral. 1st-place winner of MS COCO 2015 segmentation competition
• keywords: RoI warping layer, Multi-task Network Cascades (MNC)
• arxiv: http://arxiv.org/abs/1512.04412
• github: https://github.com/daijifeng001/MNC

Object Segmentation on SpaceNet via Multi-task Network Cascades (MNC)

• blog: https://medium.com/the-downlinq/object-segmentation-on-spacenet-via-multi-task-network-cascades-mnc-f1c89d790b42
• github: https://github.com/lncohn/pascal_to_spacenet

Learning Transferrable Knowledge for Semantic Segmentation with Deep Convolutional Neural Network

• intro: TransferNet
• project page: http://cvlab.postech.ac.kr/research/transfernet/
• arxiv: http://arxiv.org/abs/1512.07928
• github: https://github.com/maga33/TransferNet

Combining the Best of Convolutional Layers and Recurrent Layers: A Hybrid Network for Semantic Segmentation

• arxiv: http://arxiv.org/abs/1603.04871

Seed, Expand and Constrain: Three Principles for Weakly-Supervised Image Segmentation

• intro: ECCV 2016
• arxiv: https://arxiv.org/abs/1603.06098
• github: https://github.com/kolesman/SEC

ScribbleSup

ScribbleSup: Scribble-Supervised Convolutional Networks for Semantic Segmentation

• project page: http://research.microsoft.com/en-us/um/people/jifdai/downloads/scribble_sup/
• arxiv: http://arxiv.org/abs/1604.05144

Laplacian Reconstruction and Refinement for Semantic Segmentation

Laplacian Pyramid Reconstruction and Refinement for Semantic Segmentation

• intro: ECCV 2016
• arxiv: https://arxiv.org/abs/1605.02264
• paper: https://www.ics.uci.edu/~fowlkes/papers/gf-eccv16.pdf
• github(MatConvNet): https://github.com/golnazghiasi/LRR

Natural Scene Image Segmentation Based on Multi-Layer Feature Extraction

• arxiv: http://arxiv.org/abs/1605.07586

Convolutional Random Walk Networks for Semantic Image Segmentation

• arxiv: http://arxiv.org/abs/1605.07681

ENet

ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation

• arxiv: http://arxiv.org/abs/1606.02147
• github: https://github.com/e-lab/ENet-training
• github(Caffe): https://github.com/TimoSaemann/ENet
• github: https://github.com/PavlosMelissinos/enet-keras
• github: https://github.com/kwotsin/TensorFlow-ENet
• blog: http://culurciello.github.io/tech/2016/06/20/training-enet.html

Fully Convolutional Networks for Dense Semantic Labelling of High-Resolution Aerial Imagery

• arxiv: http://arxiv.org/abs/1606.02585

Deep Learning Markov Random Field for Semantic Segmentation

• arxiv: http://arxiv.org/abs/1606.07230

Region-based semantic segmentation with end-to-end training

• intro: ECCV 2016
• arxiv: http://arxiv.org/abs/1607.07671
• githun: https://github.com/nightrome/matconvnet-calvin

Built-in Foreground/Background Prior for Weakly-Supervised Semantic Segmentation

• intro: ECCV 2016
• arxiv: http://arxiv.org/abs/1609.00446

PixelNet

PixelNet: Towards a General Pixel-level Architecture

• intro: semantic segmentation, edge detection
• arxiv: http://arxiv.org/abs/1609.06694

Exploiting Depth from Single Monocular Images for Object Detection and Semantic Segmentation

• intro: IEEE T. Image Processing
• intro: propose an RGB-D semantic segmentation method which applies a multi-task training scheme: semantic label prediction and depth value regression
• arxiv: https://arxiv.org/abs/1610.01706

PixelNet: Representation of the pixels, by the pixels, and for the pixels

• intro: CMU & Adobe Research
• project page: http://www.cs.cmu.edu/~aayushb/pixelNet/
• arxiv: https://arxiv.org/abs/1702.06506
• github(Caffe): https://github.com/aayushbansal/PixelNet

Semantic Segmentation of Earth Observation Data Using Multimodal and Multi-scale Deep Networks

• arxiv: http://arxiv.org/abs/1609.06846

Deep Structured Features for Semantic Segmentation

• arxiv: http://arxiv.org/abs/1609.07916

CNN-aware Binary Map for General Semantic Segmentation

• intro: ICIP 2016 Best Paper / Student Paper Finalist
• arxiv: https://arxiv.org/abs/1609.09220

Efficient Convolutional Neural Network with Binary Quantization Layer

• arxiv: https://arxiv.org/abs/1611.06764

Mixed context networks for semantic segmentation

• intro: Hikvision Research Institute
• arxiv: https://arxiv.org/abs/1610.05854

High-Resolution Semantic Labeling with Convolutional Neural Networks

• arxiv: https://arxiv.org/abs/1611.01962

Gated Feedback Refinement Network for Dense Image Labeling

• intro: CVPR 2017
• paper: http://www.cs.umanitoba.ca/~ywang/papers/cvpr17.pdf

RefineNet

RefineNet: Multi-Path Refinement Networks with Identity Mappings for High-Resolution Semantic Segmentation

RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation

• intro: CVPR 2017. IoU 83.4% on PASCAL VOC 2012
• arxiv: https://arxiv.org/abs/1611.06612
• github: https://github.com/guosheng/refinenet
• leaderboard: http://host.robots.ox.ac.uk:8080/leaderboard/displaylb.php?challengeid=11&compid=6#KEY_Multipath-RefineNet-Res152

Full-Resolution Residual Networks for Semantic Segmentation in Street Scenes

• keywords: Full-Resolution Residual Units (FRRU), Full-Resolution Residual Networks (FRRNs)
• arxiv: https://arxiv.org/abs/1611.08323
• github(Theano/Lasagne): https://github.com/TobyPDE/FRRN
• youtube: https://www.youtube.com/watch?v=PNzQ4PNZSzc

Semantic Segmentation using Adversarial Networks

• intro: Facebook AI Research & INRIA. NIPS Workshop on Adversarial Training, Dec 2016, Barcelona, Spain
• arxiv: https://arxiv.org/abs/1611.08408
• github(Chainer): https://github.com/oyam/Semantic-Segmentation-using-Adversarial-Networks

Improving Fully Convolution Network for Semantic Segmentation

• arxiv: https://arxiv.org/abs/1611.08986

The One Hundred Layers Tiramisu: Fully Convolutional DenseNets for Semantic Segmentation

• intro: Montreal Institute for Learning Algorithms & Ecole Polytechnique de Montreal
• arxiv: https://arxiv.org/abs/1611.09326
• github: https://github.com/SimJeg/FC-DenseNet
• github: https://github.com/titu1994/Fully-Connected-DenseNets-Semantic-Segmentation
• github(Keras): https://github.com/0bserver07/One-Hundred-Layers-Tiramisu

Training Bit Fully Convolutional Network for Fast Semantic Segmentation

• intro: Megvii
• arxiv: https://arxiv.org/abs/1612.00212

Classification With an Edge: Improving Semantic Image Segmentation with Boundary Detection

• intro: “an end-to-end trainable deep convolutional neural network (DCNN) for semantic segmentation
  with built-in awareness of semantically meaningful boundaries. “
• arxiv: https://arxiv.org/abs/1612.01337

Diverse Sampling for Self-Supervised Learning of Semantic Segmentation

• arxiv: https://arxiv.org/abs/1612.01991

Mining Pixels: Weakly Supervised Semantic Segmentation Using Image Labels

• intro: Nankai University & University of Oxford & NUS
• arxiv: https://arxiv.org/abs/1612.02101

FCNs in the Wild: Pixel-level Adversarial and Constraint-based Adaptation

• arxiv: https://arxiv.org/abs/1612.02649

Understanding Convolution for Semantic Segmentation

• intro: UCSD & CMU & UIUC & TuSimple
• arxiv: https://arxiv.org/abs/1702.08502
• github(MXNet): [https://github.com/TuSimple/TuSimple-DUC]https://github.com/TuSimple/TuSimple-DUC
• pretrained-models: https://drive.google.com/drive/folders/0B72xLTlRb0SoREhISlhibFZTRmM

Label Refinement Network for Coarse-to-Fine Semantic Segmentation

https://www.arxiv.org/abs/1703.00551

Predicting Deeper into the Future of Semantic Segmentation

• intro: Facebook AI Research
• arxiv: https://arxiv.org/abs/1703.07684

Guided Perturbations: Self Corrective Behavior in Convolutional Neural Networks

• intro: University of Maryland & GE Global Research Center
• arxiv: https://arxiv.org/abs/1703.07928

Not All Pixels Are Equal: Difficulty-aware Semantic Segmentation via Deep Layer Cascade

• intro: CVPR 2017 spotlight paper
• arxxiv: https://arxiv.org/abs/1704.01344

Large Kernel Matters – Improve Semantic Segmentation by Global Convolutional Network

https://arxiv.org/abs/1703.02719

Loss Max-Pooling for Semantic Image Segmentation

• intro: CVPR 2017
• arxiv: https://arxiv.org/abs/1704.02966

Reformulating Level Sets as Deep Recurrent Neural Network Approach to Semantic Segmentation

https://arxiv.org/abs/1704.03593

A Review on Deep Learning Techniques Applied to Semantic Segmentation

https://arxiv.org/abs/1704.06857

Joint Semantic and Motion Segmentation for dynamic scenes using Deep Convolutional Networks

• intro: [International Institute of Information Technology & Max Planck Institute For Intelligent Systems
• arxiv: https://arxiv.org/abs/1704.08331

ICNet

ICNet for Real-Time Semantic Segmentation on High-Resolution Images

• intro: CUHK & Sensetime
• project page: https://hszhao.github.io/projects/icnet/
• arxiv: https://arxiv.org/abs/1704.08545
• github: https://github.com/hszhao/ICNet
• video: https://www.youtube.com/watch?v=qWl9idsCuLQ

LinkNet

Feature Forwarding: Exploiting Encoder Representations for Efficient Semantic Segmentation

LinkNet: Exploiting Encoder Representations for Efficient Semantic Segmentation

• project page: https://codeac29.github.io/projects/linknet/
• arxiv: https://arxiv.org/abs/1707.03718
• github: https://github.com/e-lab/LinkNet

Pixel Deconvolutional Networks

• intro: Washington State University
• arxiv: https://arxiv.org/abs/1705.06820

Incorporating Network Built-in Priors in Weakly-supervised Semantic Segmentation

• intro: IEEE TPAMI
• arxiv: https://arxiv.org/abs/1706.02189

Deep Semantic Segmentation for Automated Driving: Taxonomy, Roadmap and Challenges

• intro: IEEE ITSC 2017
• arxiv: https://arxiv.org/abs/1707.02432

Semantic Segmentation with Reverse Attention

• intro: BMVC 2017 oral. University of Southern California
• arxiv: https://arxiv.org/abs/1707.06426

Stacked Deconvolutional Network for Semantic Segmentation

https://arxiv.org/abs/1708.04943

Learning Dilation Factors for Semantic Segmentation of Street Scenes

• intro: GCPR 2017
• arxiv: https://arxiv.org/abs/1709.01956

A Self-aware Sampling Scheme to Efficiently Train Fully Convolutional Networks for Semantic Segmentation

https://arxiv.org/abs/1709.02764

One-Shot Learning for Semantic Segmentation

• intro: BMWC 2017
• arcxiv: https://arxiv.org/abs/1709.03410
• github: https://github.com/lzzcd001/OSLSM

An Adaptive Sampling Scheme to Efficiently Train Fully Convolutional Networks for Semantic Segmentation

https://arxiv.org/abs/1709.02764

Semantic Segmentation from Limited Training Data

https://arxiv.org/abs/1709.07665

Unsupervised Domain Adaptation for Semantic Segmentation with GANs

https://arxiv.org/abs/1711.06969

Neuron-level Selective Context Aggregation for Scene Segmentation

https://arxiv.org/abs/1711.08278

Road Extraction by Deep Residual U-Net

https://arxiv.org/abs/1711.10684

Mix-and-Match Tuning for Self-Supervised Semantic Segmentation

• intro: AAAI 2018
• project page: http://mmlab.ie.cuhk.edu.hk/projects/M&M/
• arxiv: https://arxiv.org/abs/1712.00661
• github: https://github.com/XiaohangZhan/mix-and-match/
• github: https://github.com//liuziwei7/mix-and-match

Error Correction for Dense Semantic Image Labeling

https://arxiv.org/abs/1712.03812

Semantic Segmentation via Highly Fused Convolutional Network with Multiple Soft Cost Functions

https://arxiv.org/abs/1801.01317

Instance Segmentation

Simultaneous Detection and Segmentation

• intro: ECCV 2014
• author: Bharath Hariharan, Pablo Arbelaez, Ross Girshick, Jitendra Malik
• arxiv: http://arxiv.org/abs/1407.1808
• github(Matlab): https://github.com/bharath272/sds_eccv2014

Convolutional Feature Masking for Joint Object and Stuff Segmentation

• intro: CVPR 2015
• keywords: masking layers
• arxiv: https://arxiv.org/abs/1412.1283
• paper: http://www.cv-foundation.org/openaccess/content_cvpr_2015/papers/Dai_Convolutional_Feature_Masking_2015_CVPR_paper.pdf

Proposal-free Network for Instance-level Object Segmentation

• paper: http://arxiv.org/abs/1509.02636

Hypercolumns for object segmentation and fine-grained localization

• intro: CVPR 2015
• arxiv: https://arxiv.org/abs/1411.5752
• paper: http://www.cs.berkeley.edu/~bharath2/pubs/pdfs/BharathCVPR2015.pdf

SDS using hypercolumns

• github: https://github.com/bharath272/sds

Learning to decompose for object detection and instance segmentation

• intro: ICLR 2016 Workshop
• keyword: CNN / RNN, MNIST, KITTI
• arxiv: http://arxiv.org/abs/1511.06449

Recurrent Instance Segmentation

• intro: ECCV 2016
• porject page: http://romera-paredes.com/ris
• arxiv: http://arxiv.org/abs/1511.08250
• github(Torch): https://github.com/bernard24/ris
• poster: http://www.eccv2016.org/files/posters/P-4B-46.pdf
• youtube: https://www.youtube.com/watch?v=l_WD2OWOqBk

Instance-sensitive Fully Convolutional Networks

• intro: ECCV 2016. instance segment proposal
• arxiv: http://arxiv.org/abs/1603.08678

Amodal Instance Segmentation

• intro: ECCV 2016
• arxiv: http://arxiv.org/abs/1604.08202

Bridging Category-level and Instance-level Semantic Image Segmentation

• keywords: online bootstrapping
• arxiv: http://arxiv.org/abs/1605.06885

Bottom-up Instance Segmentation using Deep Higher-Order CRFs

• intro: BMVC 2016
• arxiv: http://arxiv.org/abs/1609.02583

DeepCut: Object Segmentation from Bounding Box Annotations using Convolutional Neural Networks

• arxiv: http://arxiv.org/abs/1605.07866

End-to-End Instance Segmentation and Counting with Recurrent Attention

• intro: ReInspect
• arxiv: http://arxiv.org/abs/1605.09410

TA-FCN / FCIS

Translation-aware Fully Convolutional Instance Segmentation

Fully Convolutional Instance-aware Semantic Segmentation

• intro: CVPR 2017 Spotlight paper. winning entry of COCO segmentation challenge 2016
• arxiv: https://arxiv.org/abs/1611.07709
• github: https://github.com/msracver/FCIS
• slides: https://onedrive.live.com/?cid=f371d9563727b96f&id=F371D9563727B96F%2197213&authkey=%21AEYOyOirjIutSVk

InstanceCut: from Edges to Instances with MultiCut

• arxiv: https://arxiv.org/abs/1611.08272

Deep Watershed Transform for Instance Segmentation

• arxiv: https://arxiv.org/abs/1611.08303

Object Detection Free Instance Segmentation With Labeling Transformations

• arxiv: https://arxiv.org/abs/1611.08991

Shape-aware Instance Segmentation

• arxiv: https://arxiv.org/abs/1612.03129

Interpretable Structure-Evolving LSTM

• intro: CMU & Sun Yat-sen University & National University of Singapore & Adobe Research
• intro: CVPR 2017 spotlight paper
• arxiv: https://arxiv.org/abs/1703.03055

Mask R-CNN

• intro: ICCV 2017 Best paper award. Facebook AI Research
• arxiv: https://arxiv.org/abs/1703.06870
• github: https://github.com/TuSimple/mx-maskrcnn
• github(Keras+TensorFlow): https://github.com/matterport/Mask_RCNN

Semantic Instance Segmentation via Deep Metric Learning

https://arxiv.org/abs/1703.10277

Pose2Instance: Harnessing Keypoints for Person Instance Segmentation

https://arxiv.org/abs/1704.01152

Pixelwise Instance Segmentation with a Dynamically Instantiated Network

• intro: CVPR 2017
• arxiv: https://arxiv.org/abs/1704.02386

Instance-Level Salient Object Segmentation

• intro: CVPR 2017
• arxiv: https://arxiv.org/abs/1704.03604

Semantic Instance Segmentation with a Discriminative Loss Function

• intro: Published at “Deep Learning for Robotic Vision”, workshop at CVPR 2017. KU Leuven
• arxiv: https://arxiv.org/abs/1708.02551

SceneCut: Joint Geometric and Object Segmentation for Indoor Scenes

https://arxiv.org/abs/1709.07158

S4 Net: Single Stage Salient-Instance Segmentation

• arxiv: https://arxiv.org/abs/1711.07618
• github: https://github.com/RuochenFan/S4Net

Deep Extreme Cut: From Extreme Points to Object Segmentation

https://arxiv.org/abs/1711.09081

Learning to Segment Every Thing

• intro: UC Berkeley & Facebook AI Research
• keywords: MaskX R-CNN
• arxiv: https://arxiv.org/abs/1711.10370

Recurrent Neural Networks for Semantic Instance Segmentation

• project page: https://imatge-upc.github.io/rsis/
• arxiv: https://arxiv.org/abs/1712.00617
• github: https://github.com/imatge-upc/rsis

MaskLab

MaskLab: Instance Segmentation by Refining Object Detection with Semantic and Direction Features

https://arxiv.org/abs/1712.04837

Recurrent Pixel Embedding for Instance Grouping

• intro: learning to embed pixels and group them into boundaries, object proposals, semantic segments and instances.
• project page: http://www.ics.uci.edu/~skong2/SMMMSG.html
• arxiv: https://arxiv.org/abs/1712.08273
• github: https://github.com/aimerykong/Recurrent-Pixel-Embedding-for-Instance-Grouping
• slides: http://www.ics.uci.edu/~skong2/slides/pixel_embedding_for_grouping_public_version.pdf
• poster: http://www.ics.uci.edu/~skong2/slides/pixel_embedding_for_grouping_poster.pdf

Specific Segmentation

A CNN Cascade for Landmark Guided Semantic Part Segmentation

• project page: http://aaronsplace.co.uk/
• paper: https://aaronsplace.co.uk/papers/jackson2016guided/jackson2016guided.pdf

End-to-end semantic face segmentation with conditional random fields as convolutional, recurrent and adversarial networks

• arxiv: https://arxiv.org/abs/1703.03305

Face Parsing via Recurrent Propagation

• intro: BMVC 2017
• arxiv: https://arxiv.org/abs/1708.01936

Face Parsing via a Fully-Convolutional Continuous CRF Neural Network

https://arxiv.org/abs/1708.03736

Boundary-sensitive Network for Portrait Segmentation

https://arxiv.org/abs/1712.08675

Segment Proposal

Learning to Segment Object Candidates

• intro: Facebook AI Research (FAIR)
• intro: DeepMask. learning segmentation proposals
• arxiv: http://arxiv.org/abs/1506.06204
• github: https://github.com/facebookresearch/deepmask
• github: https://github.com/abbypa/NNProject_DeepMask

Learning to Refine Object Segments

• intro: ECCV 2016. Facebook AI Research (FAIR)
• intro: SharpMask. an extension of DeepMask which generates higher-fidelity masks using an additional top-down refinement step.
• arxiv: http://arxiv.org/abs/1603.08695
• github: https://github.com/facebookresearch/deepmask

FastMask: Segment Object Multi-scale Candidates in One Shot

• intro: CVPR 2017. University of California & Fudan University & Megvii Inc.
• arxiv: https://arxiv.org/abs/1612.08843
• github: https://github.com/voidrank/FastMask

Scene Labeling / Scene Parsing

Indoor Semantic Segmentation using depth information

• arxiv: http://arxiv.org/abs/1301.3572

Recurrent Convolutional Neural Networks for Scene Parsing

• arxiv: http://arxiv.org/abs/1306.2795
• slides: http://people.ee.duke.edu/~lcarin/Yizhe8.14.2015.pdf
• github: https://github.com/NP-coder/CLPS1520Project
• github: https://github.com/rkargon/Scene-Labeling

Learning hierarchical features for scene labeling

• paper: http://yann.lecun.com/exdb/publis/pdf/farabet-pami-13.pdf

Multi-modal unsupervised feature learning for rgb-d scene labeling

• intro: ECCV 2014
• paper: http://www3.ntu.edu.sg/home/wanggang/WangECCV2014.pdf

Scene Labeling with LSTM Recurrent Neural Networks

• paper: http://www.cv-foundation.org/openaccess/content_cvpr_2015/papers/Byeon_Scene_Labeling_With_2015_CVPR_paper.pdf

Attend, Infer, Repeat: Fast Scene Understanding with Generative Models

• arxiv: http://arxiv.org/abs/1603.08575
• notes: http://www.shortscience.org/paper?bibtexKey=journals/corr/EslamiHWTKH16

“Semantic Segmentation for Scene Understanding: Algorithms and Implementations” tutorial

• intro: 2016 Embedded Vision Summit
• youtube: https://www.youtube.com/watch?v=pQ318oCGJGY

Semantic Understanding of Scenes through the ADE20K Dataset

• arxiv: https://arxiv.org/abs/1608.05442

Learning Deep Representations for Scene Labeling with Guided Supervision

Learning Deep Representations for Scene Labeling with Semantic Context Guided Supervision

• intro: CUHK
• arxiv: https://arxiv.org/abs/1706.02493

Spatial As Deep: Spatial CNN for Traffic Scene Understanding

• intro: AAAI 2018
• arxiv: https://arxiv.org/abs/1712.06080

MPF-RNN

Multi-Path Feedback Recurrent Neural Network for Scene Parsing

• arxiv: http://arxiv.org/abs/1608.07706

Scene Labeling using Recurrent Neural Networks with Explicit Long Range Contextual Dependency

• arxiv: https://arxiv.org/abs/1611.07485

PSPNet

Pyramid Scene Parsing Network

• intro: CVPR 2017
• intro: mIoU score as 85.4% on PASCAL VOC 2012 and 80.2% on Cityscapes,
  ranked 1st place in ImageNet Scene Parsing Challenge 2016
• project page: http://appsrv.cse.cuhk.edu.hk/~hszhao/projects/pspnet/index.html
• arxiv: https://arxiv.org/abs/1612.01105
• slides: http://image-net.org/challenges/talks/2016/SenseCUSceneParsing.pdf
• github: https://github.com/hszhao/PSPNet
• github: https://github.com/Vladkryvoruchko/PSPNet-Keras-tensorflow

Open Vocabulary Scene Parsing

https://arxiv.org/abs/1703.08769

Deep Contextual Recurrent Residual Networks for Scene Labeling

https://arxiv.org/abs/1704.03594

Fast Scene Understanding for Autonomous Driving

• intro: Published at “Deep Learning for Vehicle Perception”, workshop at the IEEE Symposium on Intelligent Vehicles 2017
• arxiv: https://arxiv.org/abs/1708.02550

FoveaNet: Perspective-aware Urban Scene Parsing

https://arxiv.org/abs/1708.02421

BlitzNet: A Real-Time Deep Network for Scene Understanding

• intro: INRIA
• arxiv: https://arxiv.org/abs/1708.02813

Semantic Foggy Scene Understanding with Synthetic Data

https://arxiv.org/abs/1708.07819

Restricted Deformable Convolution based Road Scene Semantic Segmentation Using Surround View Cameras

https://arxiv.org/abs/1801.00708

Benchmarks

MIT Scene Parsing Benchmark

• homepage: http://sceneparsing.csail.mit.edu/
• github(devkit): https://github.com/CSAILVision/sceneparsing

Semantic Understanding of Urban Street Scenes: Benchmark Suite

https://www.cityscapes-dataset.com/benchmarks/

Challenges

Large-scale Scene Understanding Challenge

• homepage: http://lsun.cs.princeton.edu/

Places2 Challenge

http://places2.csail.mit.edu/challenge.html

Human Parsing

Human Parsing with Contextualized Convolutional Neural Network

• intro: ICCV 2015
• paper: http://www.cv-foundation.org/openaccess/content_iccv_2015/html/Liang_Human_Parsing_With_ICCV_2015_paper.html

Look into Person: Self-supervised Structure-sensitive Learning and A New Benchmark for Human Parsing

• intro: CVPr 2017. SYSU & CMU
• keywords: Look Into Person (LIP)
• project page: http://hcp.sysu.edu.cn/lip/
• arxiv: https://arxiv.org/abs/1703.05446
• github: https://github.com/Engineering-Course/LIP_SSL

Cross-domain Human Parsing via Adversarial Feature and Label Adaptation

• intro: AAAI 2018
• arxiv: https://arxiv.org/abs/1801.01260

Video Object Segmentation

Fast object segmentation in unconstrained video

• project page: http://calvin.inf.ed.ac.uk/software/fast-video-segmentation/
• paper: http://calvin.inf.ed.ac.uk/wp-content/uploads/Publications/papazoglouICCV2013-camera-ready.pdf

Recurrent Fully Convolutional Networks for Video Segmentation

• arxiv: https://arxiv.org/abs/1606.00487

Object Detection, Tracking, and Motion Segmentation for Object-level Video Segmentation

• arxiv: http://arxiv.org/abs/1608.03066

Clockwork Convnets for Video Semantic Segmentation

• intro: ECCV 2016 Workshops
• intro: evaluated on the Youtube-Objects, NYUD, and Cityscapes video datasets
• arxiv: http://arxiv.org/abs/1608.03609
• github: https://github.com/shelhamer/clockwork-fcn

STFCN: Spatio-Temporal FCN for Semantic Video Segmentation

• arxiv: http://arxiv.org/abs/1608.05971

One-Shot Video Object Segmentation

• intro: OSVOS
• project: http://www.vision.ee.ethz.ch/~cvlsegmentation/osvos/
• arxiv: https://arxiv.org/abs/1611.05198
• github: https://github.com/kmaninis/OSVOS-caffe
• github: https://github.com/scaelles/OSVOS-TensorFlow

Video Object Segmentation Without Temporal Information

https://arxiv.org/abs/1709.06031

Convolutional Gated Recurrent Networks for Video Segmentation

• arxiv: https://arxiv.org/abs/1611.05435

Learning Video Object Segmentation from Static Images

• arxiv: https://arxiv.org/abs/1612.02646

Semantic Video Segmentation by Gated Recurrent Flow Propagation

• arxiv: https://arxiv.org/abs/1612.08871

FusionSeg: Learning to combine motion and appearance for fully automatic segmention of generic objects in videos

• project page: http://vision.cs.utexas.edu/projects/fusionseg/
• arxiv: https://arxiv.org/abs/1701.05384
• github: https://github.com/suyogduttjain/fusionseg

Unsupervised learning from video to detect foreground objects in single images

https://arxiv.org/abs/1703.10901

Semantically-Guided Video Object Segmentation

https://arxiv.org/abs/1704.01926

Learning Video Object Segmentation with Visual Memory

https://arxiv.org/abs/1704.05737

Flow-free Video Object Segmentation

https://arxiv.org/abs/1706.09544

Online Adaptation of Convolutional Neural Networks for Video Object Segmentation

https://arxiv.org/abs/1706.09364

Video Object Segmentation using Tracked Object Proposals

• intro: CVPR-2017 workshop, DAVIS-2017 Challenge
• arxiv: https://arxiv.org/abs/1707.06545

Video Object Segmentation with Re-identification

• intro: CVPR 2017 Workshop, DAVIS Challenge on Video Object Segmentation 2017 (Winning Entry)
• arxiv: https://arxiv.org/abs/1708.00197

Pixel-Level Matching for Video Object Segmentation using Convolutional Neural Networks

• intro: ICCV 2017
• arxiv: https://arxiv.org/abs/1708.05137

SegFlow: Joint Learning for Video Object Segmentation and Optical Flow

• project page: https://sites.google.com/site/yihsuantsai/research/iccv17-segflow
• arxiv: https://arxiv.org/abs/1709.06750
• github: https://github.com/JingchunCheng/SegFlow

Video Semantic Object Segmentation by Self-Adaptation of DCNN

https://arxiv.org/abs/1711.08180

Learning to Segment Moving Objects

https://arxiv.org/abs/1712.01127

Instance Embedding Transfer to Unsupervised Video Object Segmentation

• intro: University of Southern California & Google Inc
• arxiv: https://arxiv.org/abs/1801.00908

Panoptic Segmentation

• intro: Facebook AI Research (FAIR) & Heidelberg University
• arxiv: https://arxiv.org/abs/1801.00868

Challenge

DAVIS: Densely Annotated VIdeo Segmentation

• homepage: http://davischallenge.org/
• arxiv: https://arxiv.org/abs/1704.00675

DAVIS Challenge on Video Object Segmentation 2017

http://davischallenge.org/challenge2017/publications.html

Projects

TF Image Segmentation: Image Segmentation framework

• intro: Image Segmentation framework based on Tensorflow and TF-Slim library
• github: https://github.com/warmspringwinds/tf-image-segmentation

KittiSeg: A Kitti Road Segmentation model implemented in tensorflow.

• keywords: MultiNet
• intro: KittiSeg performs segmentation of roads by utilizing an FCN based model.
• github: https://github.com/MarvinTeichmann/KittiBox

Semantic Segmentation Architectures Implemented in PyTorch

• intro: Segnet/FCN/U-Net/Link-Net
• github: https://github.com/meetshah1995/pytorch-semseg

PyTorch for Semantic Segmentation

https://github.com/ZijunDeng/pytorch-semantic-segmentation

3D Segmentation

PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation

• intro: Stanford University
• project page: http://stanford.edu/~rqi/pointnet/
• arxiv: https://arxiv.org/abs/1612.00593
• github: https://github.com/charlesq34/pointnet

DA-RNN: Semantic Mapping with Data Associated Recurrent Neural Networks

https://arxiv.org/abs/1703.03098

SqueezeSeg: Convolutional Neural Nets with Recurrent CRF for Real-Time Road-Object Segmentation from 3D LiDAR Point Cloud

• intro: UC Berkeley
• arxiv: https://arxiv.org/abs/1710.07368

SEGCloud: Semantic Segmentation of 3D Point Clouds

• intro: International Conference of 3D Vision (3DV) 2017 (Spotlight). Stanford University
• homepage: http://segcloud.stanford.edu/
• arxiv: https://arxiv.org/abs/1710.07563

Leaderboard

Segmentation Results: VOC2012 BETA: Competition “comp6” (train on own data)

http://host.robots.ox.ac.uk:8080/leaderboard/displaylb.php?cls=mean&challengeid=11&compid=6

Blogs

Deep Learning for Natural Image Segmentation Priors

http://cs.brown.edu/courses/csci2951-t/finals/ghope/

Image Segmentation Using DIGITS 5

https://devblogs.nvidia.com/parallelforall/image-segmentation-using-digits-5/

Image Segmentation with Tensorflow using CNNs and Conditional Random Fields
http://warmspringwinds.github.io/tensorflow/tf-slim/2016/12/18/image-segmentation-with-tensorflow-using-cnns-and-conditional-random-fields/

Fully Convolutional Networks (FCNs) for Image Segmentation

• blog: http://warmspringwinds.github.io/tensorflow/tf-slim/2017/01/23/fully-convolutional-networks-(fcns)-for-image-segmentation/
• ipn: https://github.com/warmspringwinds/tensorflow_notes/blob/master/fully_convolutional_networks.ipynb

Image segmentation with Neural Net

• blog: https://medium.com/@m.zaradzki/image-segmentation-with-neural-net-d5094d571b1e#.s5f711g1q
• github: https://github.com/mzaradzki/neuralnets/tree/master/vgg_segmentation_keras

A 2017 Guide to Semantic Segmentation with Deep Learning

http://blog.qure.ai/notes/semantic-segmentation-deep-learning-review

Talks

Deep learning for image segmentation

• intro: PyData Warsaw - Mateusz Opala & Michał Jamroż
• youtube: https://www.youtube.com/watch?v=W6r_a5crqGI