본문 바로가기

Awesome

Awesome Road Environment Segmentation

728x90
반응형

Captured from Cityscapes (https://www.cityscapes-dataset.com/examples/)

Networks by architecture

DeepLabv3+ paper

HRNetV2 paper

Datasets

Name Categories Volume Link
Cityscapes road, sidewalk, guard rail, ... 30 classes / 25,000 images Features
Mapillary (Vistas) road, curb, barrier, wall, ... 152 classes / 25,000 images Features
A2D21) road, sidebars, curbstone, ... 38 classes / 41,280 images Features
ApolloScape (Scene Parsing) road, sidewalk, fence, ... 37 classes / 74,555 images Features
DeepDrive2) road, sidewalk, guard rail, ... 40 classes / 10,000 images Features
KITTI2) road, sidewalk, guard rail, ... 34 classes / 400 images Features

1) A2D2: AEV(Audi Electronics Venture) Autonomous Driving Dataset

2) Class code inspired by 'cityscapes'. (ref.)

Annotation tools

Getting Started Guide

Benchmarks

Evaluation of benchmarks

Papers, blog posts and others

Papers

Blog posts

Awesome series

Reference repositories

Others

 


https://github.com/DaveLogs/awesome-road-environment-segmentation

 

DaveLogs/awesome-road-environment-segmentation

Awesome lists about segmentation in road environment 😁 - DaveLogs/awesome-road-environment-segmentation

github.com

 

 

728x90
반응형

'Awesome' 카테고리의 다른 글

Awesome Powerful and Useful Ubuntu  (0) 2021.06.01