Explore and analyze large datasets of images
Project description
Panoptic 👀
This documentation also exist in French
Panoptic is a tool for exploring and annotating large image corpora, using image analysis and machine learning tools to facilitate these tasks.
Since it requires deep learning libraries, it is recommended to use it with a computer with minimal computing capabilities.
Caution: Panoptic is still in active development and is currently a prototype, it is likely that you will encounter bugs, so we recommend to use this tool only for testing and not to rely on it for a substantial academic work.
Windows and Linux installation
Python 3.9 or higher is required, and a custom environnment (venv, pyenv, conda etc.) is recommended. Open a terminal and type:
pip install panoptic
panoptic
Mac installation
Just like the windows installation but you will need to also install the xcode tools first.
For this, just use this command in a terminal: xcode-select –-install
, it should trigger the installation of the command line tools.
Then run:
pip install panoptic
panoptic
Docker installation
If you're having issues with the installation (sometimes the packages can be tricky) or if you just prefair to use Docker it's possible: Start by installing docker:
Option 1 : One folder to rule them all
Make a folder for the docker with a subfolder for the images /path/to/your/folder/images
.
docker run -it -p 8000:8000 -v /path/to/your/folder:/data --name panoptic ceressorbonne/panoptic
Option 2 : Separate data and images folders
Make a folder for the panoptic datas (database, thumbnails, etc.) /path/to/your/data/
.
And another one withe your images (could be an existing one) /path/to/your/images/
.
docker run -it -p 8000:8000 \
-v /path/to/your/data:/data \
-v /path/to/your/images:/data/images \
--name panoptic \
ceressorbonne/panoptic
Access
You can now access panoptic through: https://localhost:8000 As if you were to launch it with the python version
Restart
If you happen to stop panoptic, you can re launch it with :
docker start -ia panoptic
Installation (development)
The following steps involve cloning the directory and are recommended for users who wish to have access to the development versions, or who wish to modify the code themselves in order to contribute.
Backend development only
To test and modify the backend operation, we provide an already built frontend in the backend html folder.
- go to the
panoptic-back
folder - to install the dependencies
python setup.py install
to simply use panopticpip install -e .
to developpip install -r requirements.txt
and you have to addpanoptic-back
to the PYTHON_PATH for developping purpose too
- run
python panoptic/main.py
Front and back development
- First of all, perform the installation steps of the backend
- go to the
panoptic-front
folder - Run
npm install
. - run
npm run dev
. - before running the backend the
PANOPTIC_ENV
environment variable should be set toDEV
in order to use the development frontend.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for panoptic-0.3.5rc10-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 7967e256044287cde864f1bf820134c1e6d63b2399ffc346b90b2cae3881c4ce |
|
MD5 | 46134417fe3f1b8ac0ac0efba25f57d2 |
|
BLAKE2b-256 | a32bcb7467798093a1b7ff517f0f3512e894e14fb48f7d37e7c722f0d9bd9e39 |