GithubHelp home page GithubHelp logo

sar_specific_models's Introduction

🆕 The Latest Version of DSN https://github.com/Alien9427/DSN

@article{dsn2020,
title = {Deep SAR-Net: Learning objects from signals},
journal = {ISPRS Journal of Photogrammetry and Remote Sensing},
volume = {161},
pages = {179-193},
year = {2020},
issn = {0924-2716},
author = {Z. Huang and M. Datcu and Z. Pan and B. Lei},
}

🆕 More SAR image pre-trained models are referred to SAR-HUB

SAR_specific_models

This project provides some SAR specific models with strong abilities to extract spatial features of single-polarization Synthetic Aperture Radar (SAR) amplitude images.

Environment

Pytorch 0.4.0 (also verified in Pytorch 1.4.0) Python 3.6

SAR-Specific Models

./model/resnet18_I_nwpu_tsx.pth [1]

The SAR image pre-trained model in Reference [1].

It can be transferred to other SAR classification or detection models with ResNet-18 backbone.

Transfer to Target Detection

Use MMDetection to transfer the pretrained model to SAR detection:

Picture1

import torchvision.models as models
resnet18 = models.resnet18(pretrained=False)
pthfile = './model/resnet18_I_nwpu_tsx.pth'
resnet18.load_state_dict(torch.load(pthfile))

./model/resnet18_tsx_mstar_epoch7.pth [1]

The transferred model to MSTAR 10-class target recognition task in Reference [1], achieving an overall accuracy of 99.46%.

./model/alexnet_tsx.pth [2]

The SAR-image pre-trained model in Reference [2].

./model/alexnet_tsx_mstar_iter1920.pth [2]

The transferred model to MSTAR 10-class target recognition task in Reference [2], achieving an overall accuracy of 99.34%.

References

[1] Classification of Large-Scale High-Resolution SAR Images with Deep Transfer Learning, IEEE GRSL 2020

doi: 10.1109/LGRS.2020.2965558

[2] What, Where and How to Transfer in SAR Target Recognition Based on Deep CNNs, IEEE TGRS 2019

doi: 10.1109/TGRS.2019.2947634

sar_specific_models's People

Contributors

alien9427 avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

sar_specific_models's Issues

关于权重文件

是只有./model/resnet18_I_nwpu_tsx.pth这个模型可以用于mmdetection的目标检测吗?通过joint训练出来的模型可以用嘛?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.