GitPlanet

Cheap and reliable Node.js hosting starts at $3/month, and $1/month static HTML hosting

Created with love in Canada, visit hostnodejs.com today

Feel like to post an Ad? Learn Details
All Projects → LittleWat → multichannel-semseg-with-uda

LittleWat / multichannel-semseg-with-uda

Licence: other
Multichannel Semantic Segmentation with Unsupervised Domain Adaptation

Programming Languages

python
139335 projects - #7 most used programming language
shell
77523 projects
matlab
3953 projects

Labels

pytorchmultichannelsemantic-segmentationdomain-adaptation

Projects that are alternatives of or similar to multichannel-semseg-with-uda

AdaptationSeg
Curriculum Domain Adaptation for Semantic Segmentation of Urban Scenes, ICCV 2017
Stars: ✭ 128 (+573.68%)
Mutual labels:  semantic-segmentation, domain-adaptation
Adaptsegnet
Learning to Adapt Structured Output Space for Semantic Segmentation, CVPR 2018 (spotlight)
Stars: ✭ 654 (+3342.11%)
Mutual labels:  semantic-segmentation, domain-adaptation
IAST-ECCV2020
IAST: Instance Adaptive Self-training for Unsupervised Domain Adaptation (ECCV 2020) https://teacher.bupt.edu.cn/zhuchuang/en/index.htm
Stars: ✭ 84 (+342.11%)
Mutual labels:  semantic-segmentation, domain-adaptation
Seanet
Self-Ensembling Attention Networks: Addressing Domain Shift for Semantic Segmentation
Stars: ✭ 90 (+373.68%)
Mutual labels:  semantic-segmentation, domain-adaptation
Cbst
Code for <Domain Adaptation for Semantic Segmentation via Class-Balanced Self-Training> in ECCV18
Stars: ✭ 146 (+668.42%)
Mutual labels:  semantic-segmentation, domain-adaptation
Intrada
Unsupervised Intra-domain Adaptation for Semantic Segmentation through Self-Supervision (CVPR 2020 Oral)
Stars: ✭ 211 (+1010.53%)
Mutual labels:  semantic-segmentation, domain-adaptation
LoveDA
[NeurIPS2021 Poster] LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation
Stars: ✭ 111 (+484.21%)
Mutual labels:  semantic-segmentation, domain-adaptation
Lsd Seg
Learning from Synthetic Data: Addressing Domain Shift for Semantic Segmentation
Stars: ✭ 99 (+421.05%)
Mutual labels:  semantic-segmentation, domain-adaptation
Dise Domain Invariant Structure Extraction
Pytorch Implementation -- All about Structure: Adapting Structural Information across Domains for Boosting Semantic Segmentation, CVPR 2019
Stars: ✭ 129 (+578.95%)
Mutual labels:  semantic-segmentation, domain-adaptation
Seg Uncertainty
IJCAI2020 & IJCV 2020 🌇 Unsupervised Scene Adaptation with Memory Regularization in vivo
Stars: ✭ 202 (+963.16%)
Mutual labels:  semantic-segmentation, domain-adaptation
Clan
( CVPR2019 Oral ) Taking A Closer Look at Domain Shift: Category-level Adversaries for Semantics Consistent Domain Adaptation
Stars: ✭ 248 (+1205.26%)
Mutual labels:  semantic-segmentation, domain-adaptation
UniFormer
[ICLR2022] official implementation of UniFormer
Stars: ✭ 574 (+2921.05%)
Mutual labels:  semantic-segmentation
CSSR
Crack Segmentation for Low-Resolution Images using Joint Learning with Super-Resolution (CSSR) was accepted to international conference on MVA2021 (oral), and selected for the Best Practical Paper Award.
Stars: ✭ 50 (+163.16%)
Mutual labels:  semantic-segmentation
Deep-Unsupervised-Domain-Adaptation
Pytorch implementation of four neural network based domain adaptation techniques: DeepCORAL, DDC, CDAN and CDAN+E. Evaluated on benchmark dataset Office31.
Stars: ✭ 50 (+163.16%)
Mutual labels:  domain-adaptation
Fashion-Clothing-Parsing
FCN, U-Net models implementation in TensorFlow for fashion clothing parsing
Stars: ✭ 29 (+52.63%)
Mutual labels:  semantic-segmentation
pytorch-UNet
2D and 3D UNet implementation in PyTorch.
Stars: ✭ 107 (+463.16%)
Mutual labels:  semantic-segmentation
PSPNet-Pytorch
Implemetation of Pyramid Scene Parsing Network in Pytorch
Stars: ✭ 26 (+36.84%)
Mutual labels:  semantic-segmentation
StyleGAN-nada
stylegan-nada.github.io/
Stars: ✭ 1,018 (+5257.89%)
Mutual labels:  domain-adaptation
SEC-tensorflow
a tensorflow version for SEC approach in the paper "seed, expand and constrain: three principles for weakly-supervised image segmentation".
Stars: ✭ 35 (+84.21%)
Mutual labels:  semantic-segmentation
pysemseg
Semantic Segmentation Models in Pytorch
Stars: ✭ 31 (+63.16%)
Mutual labels:  semantic-segmentation
View All Similar Projects ➔

Multichannel Semantic Segmentation with Unsupervised Domain Adaptation implemeted by PyTorch

This is the code for the paper (Multichannel Semantic Segmentation with Unsupervised Domain Adaptation) in AutoNUE workshop at ECCV-2018.

Setting

setting

Sample results

result

If you find this code(sorry still too messy) useful in your research, please consider citing:

@inproceedings{watanabe2018multichannel,
  title={Multichannel Semantic Segmentation with Unsupervised Domain Adaptation},
  author={Watanabe, Kohei and Saito, Kuniaki and Ushiku, Yoshitaka and Harada, Tatsuya},
  booktitle={Proceedings of the on AUTONUE Workshops of ECCV 2018},
  year={2018},
  organization={Springer}
}

Installation

Use Python 2.x

First, you need to install PyTorch following the official site instruction.

Next, please install the required libraries as follows;

pip install -r requirements.txt

Dataset Preparation

Please download datasets from URLs below;

Then, edit the get_dataset function in datasets.py.

Demo

demo

You can try our demo online

First, download the trained model as follows;

wget https://www.dropbox.com/s/4lis0cjju5ounlg/dual_model.tar

Then, run the demo script as follows;

python demo.py  sample_img/rgb_5947.png dual_model.tar

Result will be saved under demo_output directory.

Usage

We adopted Maximum Classifier Discrepancy (MCD) for unsupervised domain adaptation.

MCD Training

  • adapt_xxx.py

    • for domain adaptation (MCD)

  • dann_xxx.py

    • for domain adaptation (DANN: Domain Adversarial Neural Network)

  • source_xxx.py

    • for source only

Fusion-based approach

Early Fusion

python adapt_trainer.py suncg nyu --input_ch 6 --src_split train_rgbhhab --tgt_split trainval_rgbhha

Late Fusion

python adapt_mfnet_trainer.py suncg nyu --input_ch 6 --src_split train_rgbhhab --tgt_split trainval_rgbhha --method_detail MFNet-AddFusion

Score Fusion

python adapt_mfnet_trainer.py suncg nyu --input_ch 6 --src_split train_rgbhhab --tgt_split trainval_rgbhha --method_detail MFNet-ScoreAddFusion

Multitask learning approach

Segmentation + Depth Estimation (HHA regression)

python adapt_multitask_trainer.py suncg nyu --input_ch 6 --src_split train_rgbhhab --tgt_split trainval_rgbhha --method_detail MFNet-ScoreAddFusion

Segmentation + Depth Estimation (HHA regression) + Boundary Detection

python adapt_tripletask_trainer.py suncg nyu --input_ch 6 --src_split train_rgbhhab --tgt_split trainval_rgbhha  --method_detail MFNet-ScoreAddFusion

Test

For dual task,

python adapt_multitask_tester.py nyu --split test_rgbhha train_output/suncg-train_rgbhha2nyu-trainval_rgbhha_6ch_MCDmultitask/pth/MCD-normal-drn_d_38-20.pth.tar

For triple task,

python adapt_triple_multitask_tester.py nyu --split test_rgbhha train_output/suncg-train_rgbhhab2nyu-trainval_rgbhha_6ch_MCD_triple_multitask/pth/MCD-normal-drn_d_38-10.pth.tar

Results will be saved under "./test_output/suncg-train_rgbhhab2nyu-trainval_rgbhha_6ch_MCD_triple_multitask---nyu-test_rgbhha/MCD-normal-drn_d_38-10.tar/" .

Postprocess using Boundary Detection output

You need Matlab.

bash ./sample_scripts/refine_seg_by_boundary.sh

Evaluation

python eval.py nyu ./test_output/suncg-train_rgbhhab2nyu-trainval_rgbhha_6ch_MCD_triple_multitask---nyu-test_rgbhha/YOUR_MODEL_NAME/label

Referenced codes

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].