https://github.com/khdlr/soundingearth
Self-supervised Audiovisual Representation Learning for Remote Sensing Data
https://github.com/khdlr/soundingearth
Last synced: over 1 year ago
JSON representation
Self-supervised Audiovisual Representation Learning for Remote Sensing Data
- Host: GitHub
- URL: https://github.com/khdlr/soundingearth
- Owner: khdlr
- Created: 2021-07-19T09:36:15.000Z (almost 5 years ago)
- Default Branch: master
- Last Pushed: 2023-05-22T07:44:38.000Z (about 3 years ago)
- Last Synced: 2024-04-12T07:18:26.980Z (over 2 years ago)
- Language: Python
- Homepage:
- Size: 66.4 KB
- Stars: 25
- Watchers: 3
- Forks: 4
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# SoundingEarth
Code for [Self-supervised Audiovisual Representation Learning for Remote Sensing Data](https://arxiv.org/abs/2108.00688)
## Data
Download the dataset at https://zenodo.org/record/5600379 and configure `DataRoot` in `config.yml` to point to the extracted archive.
## Code
* `train.py` is the main engine for the project.
* Configuration is done in `config.py` (defaults) and `config.yml` (experiments)
* Data loading in `data_loading.py`
All other interesting stuff goes in `lib/`:
* Models to `lib/models/`
* Loss functions in `lib/loss_functions.py`
## Pre-trained Weights
Find model weights in the [Releases](https://github.com/khdlr/SoundingEarth/releases) tab