【语义分割】Fast-SCNN|【语义分割】Fast-SCNN -- Fast Semantic Segmentation Network 【语义分割】Fast-SCNN--FastSema

paper：Fast-SCNN
github：code

文章目录

Abstract
Fast-SCNN

Network Architecture
Pre-training on Auxiliary Tasks

Experiment

1、Evaluation on Cityscapes
2、Pre-training and Weakly Labeled Data
3、Lower Input Resolution

Abstract 【【语义分割】Fast-SCNN|【语义分割】Fast-SCNN -- Fast Semantic Segmentation Network】Fast-SCNN是一个实时的语义分割模型。其基于现有的two-branch方法（BiSeNet），引入了一个learning to downsample模块，在cityscapes上得到68.0%的miou。
FastSCNN采用depthwise separable convolutions和inverse residual blocks
Fast-SCNN 实时语义分割模型设计要点：

a larger receptive field is important to learn complex correlations among object classes (i.e. global context)
spatial detail in images is necessary to preserve object
boundaries
balance speed and accuracy

Network Architecture

【语义分割】Fast-SCNN|【语义分割】Fast-SCNN -- Fast Semantic Segmentation Network

文章图片

文章图片
PPM是PSPNet中的一个重要模块，可融合 different-region-based context information.

文章图片
Adding few layers after the feature fusion module boosts the accuracy
Pre-training on Auxiliary Tasks In our experiments we show that small networks do not get significant benefit from pre-training. Instead, aggressive data augmentation and more number of epochs provide similar results
Fast-SCNN在cityscapes上训练1000个epoch
Experiment 1、Evaluation on Cityscapes