Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/georgia-tech-db/eva-labeling
Labeling application on top of EVA
https://github.com/georgia-tech-db/eva-labeling
Last synced: about 2 months ago
JSON representation
Labeling application on top of EVA
- Host: GitHub
- URL: https://github.com/georgia-tech-db/eva-labeling
- Owner: georgia-tech-db
- License: apache-2.0
- Created: 2022-10-05T14:30:28.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2023-04-13T16:43:08.000Z (over 1 year ago)
- Last Synced: 2024-04-24T11:13:50.936Z (8 months ago)
- Language: Python
- Size: 3 MB
- Stars: 7
- Watchers: 11
- Forks: 4
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# eva-labeling
## What is EVA Labeling?
EVA Labeling is a wrapper to enable you to run EVA alongside [Label-Studio](!https://labelstud.io/) It lets you connect to Label Studio server to do the following:
- Annotate your dataset more quickly and easily by allowing "bulk labeling" of a set of images.
- Dynamically pre-annotate data based on model inference results.
- Retrain or fine-tune a model based on recently annotated data.
## Demo video
https://user-images.githubusercontent.com/57455619/229378952-b1048139-3d05-4377-a1bb-83a39fc174d6.mp4
## How it works
1. It first fetches pairs from Label Studio server and loads them into the EVA database server.
2. It then runs queries over the loaded data within the EVA database server (e.g., image classification query, object detection query)
3. Finally, it sends the query results (i.e., the updated labels) back to the Label Studio server.
## Quickstart
> :warning: The Label Studio server should be started with flag **EXPERIMENTAL_FEATURES=1** for the "Bulk Label Propagation" feature to work!
1. Install the dependencies in a virtual environment
```bash
# Install dependencies
pip install -r requirements.txt
```
2. Start EVA Labeling Server
> Note: This command will automatically start the `EVA` server.
```bash
label-studio-ml start ./evaml -eu -ep -k -ls
```
3. Add the Image Clustering Interface
```html
```
4. Start and Register the EVA ML Backend to Label Studio.
> Following [Label Studio ML documentation](https://github.com/heartexlabs/label-studio-ml-backend)
## Adding Custom Models to EVA
1. Register your Feature Extractor model into EVA.
> Refer the [EVA documentation](https://evadb.readthedocs.io/en/stable/source/reference/udf.html) for more information.
> [Here](https://evadb.readthedocs.io/en/stable/source/tutorials/04-custom-model.html) is an example of adding a custom model.
2. You can add your custom queries inside `./evaml/cluster_image.py`,
> Refer the [Label Studio ML Backend documentation](https://github.com/heartexlabs/label-studio-ml-backend) for more information.