https://github.com/isaaccorley/contrastive-surface-image-pretraining

The official repository for the ICIP 2022 paper "Supervising Remote Sensing Change Detection Models with 3D Surface Semantics"
https://github.com/isaaccorley/contrastive-surface-image-pretraining

Last synced: 3 months ago
JSON representation

The official repository for the ICIP 2022 paper "Supervising Remote Sensing Change Detection Models with 3D Surface Semantics"

• Host: GitHub
• URL: https://github.com/isaaccorley/contrastive-surface-image-pretraining
• Owner: isaaccorley
• License: mit
• Created: 2022-02-26T18:35:06.000Z (almost 3 years ago)
• Default Branch: main
• Last Pushed: 2023-11-28T17:55:48.000Z (about 1 year ago)
• Last Synced: 2024-05-02T05:05:17.312Z (8 months ago)
• Language: Python
• Homepage:
• Size: 38.3 MB
• Stars: 6
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Contrastive Surface Image Pretraining (CSIP)



This repository contains code for the ICIP 2022 paper ["Supervising Remote Sensing Change Detection Models with 3D Surface Semantics", Corley et al. (2022)](https://arxiv.org/abs/2202.13251). In this paper we contrastively train dual encoders (RGB and AGL) to learn 3D surface semantics. We then evaluate the pretrained RGB encoder on several downstream change detection and segmentation tasks. We find that pretraining using our CSIP framework improves performance on downstream tasks where surface and height information is relevant in optical imagery.



## Install dependencies

First install OpenJPEG to read JPEG2000 (.j2k) files

```bash

sudo apt-get install -y libopenjp2-7 libopenjp2-7-dev

```

Install Python dependencies

```bash

pip install -r requirements.txt

```

## Datasets

- The Overhead Geopose dataset can be downloaded [here](https://ieee-dataport.org/open-access/urban-semantic-3d-dataset)

- The OSCD dataset can be downloaded [here](https://rcdaudt.github.io/oscd/)

- The xBD dataset can be downloaded [here](https://xview2.org/dataset)

- The S2Looking dataset can be downloaded [here](https://github.com/S2Looking/Dataset)

- The LandCover.ai dataset can be downloaded [here](https://landcover.ai/)

## Pretraining

```bash

python pretrain.py --cfg conf/pretrain_overhead_geopose.yaml

```

## Training

To perform an individual experiment, use the following:

```bash

python train.py --cfg conf/oscd-csip.yaml

```

To repeat all experiments from the paper, use the following:

```bash

bash train.sh

```

## Evaluation

To evaluate an individual experiment, use the following:

```bash

python evaluate.py --dir logs/oscd-csip

```

To evaluate all experiments from the paper, use the following:

```bash

bash evaluate.sh

```