A 2017 Guide to Semantic Segmentation with Deep Learning 笔记
2018-03-23 17:26
501 查看
原文A 2017 Guide to Semantic Segmentation with Deep Learning
0. Intro
1. Problem
1.1 before deep
1.2 current
1.3 postprocessing
2. Models
why medical images are different from natural images
dataset: VOC2012, MSCOCO
metric: mean IOU
before deep
textonforest
random forest based classifier
prob
classifaction fixed input size
pooling layer: discard ‘where’ infomation
patch classication
classification networks usually have full connected layers and therefore required fixed size images.
1.2 current
FCN(prob1)
allow segmentation on any size image
Pooling(prob2)
encoder-decoder arch
dilated conv
encoder-decoder
encoder: reduces the spatial dimension with pooling layer
decoder: recover object details and spatial dimension
shortcut connections: help decoder recover the object details better
dilated conv
away with pooling layers
1.3 postprocessing
CRF postprocessing
similar intensity pixels tend to be labeled as same class
boost scores by 1-2%
0. Intro
1. Problem
1.1 before deep
1.2 current
1.3 postprocessing
2. Models
0. Intro
mainly use natural/real world image datasetswhy medical images are different from natural images
dataset: VOC2012, MSCOCO
metric: mean IOU
1. Problem
1.1 before deepbefore deep
textonforest
random forest based classifier
prob
classifaction fixed input size
pooling layer: discard ‘where’ infomation
patch classication
classification networks usually have full connected layers and therefore required fixed size images.
1.2 current
FCN(prob1)
allow segmentation on any size image
Pooling(prob2)
encoder-decoder arch
dilated conv
encoder-decoder
encoder: reduces the spatial dimension with pooling layer
decoder: recover object details and spatial dimension
shortcut connections: help decoder recover the object details better
dilated conv
away with pooling layers
1.3 postprocessing
CRF postprocessing
similar intensity pixels tend to be labeled as same class
boost scores by 1-2%
2. Models
相关文章推荐
- A 2017 Guide to Semantic Segmentation with Deep Learning
- 论文笔记:A Review on Deep Learning Techniques Applied to Semantic Segmentation
- A guide to convolution arithmetic for deep learning学习笔记
- 【Review】A Review on Deep Learning Techniques Applied to Semantic Segmentation
- Learning to Track at 100 FPS with Deep Regression Networks 论文理解及应用笔记(一)
- Learning to Track at 100 FPS with Deep Regression Networks 论文理解及应用笔记(二)
- Learning to Track at 100 FPS with Deep Regression Networks 论文笔记
- A Review on Deep Learning Techniques Applied to Semantic Segmentation(译)-(1)
- A Review on Deep Learning Techniques Applied to Semantic Segmentation(译)-(2)
- 【学习笔记】WEEK 1_Introduction to Deep Learning_Supervised Learning with Neural Networks
- DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution--阅读笔记
- A Guide to Deep Learning
- [深度学习论文笔记][Weight Initialization] Exact solutions to the nonlinear dynamics of learning in deep lin
- Learning Deep Structured Semantic Models for Web Search using Clickthrough Data笔记
- Introduction.to.Machine.Learning.with.Python 笔记
- 生成对抗网络学习笔记5----DCGAN(unsupervised representation learning with deep convolutional generative adv)的实现
- 【学习笔记】WEEK 1_Introduction to Deep Learning_Why is Deep Learning taking off?
- 深度学习Tracking(1)——Learning to Track at 100 FPS with Deep Regression Networks
- ISSCC 2017论文导读 Session 14:A 288μW Programmable Deep-Learning Processor with 270KB On-Chip Weight
- 【论文阅读笔记】CheXNet: Radiologist-Level Pneumonia Detection on Chest X-Rays with Deep Learning