This is a SegServe executor for the mmdetection segmentation and detection framework supporting a wide variety of machine learing approaches.
Make sure you have anaconda installed and an active environment with mlflow
. Then execute
pip install mlflow
mlflow run ./ -e main -P input_images=<path to your local image or image folder (*.png)> -P config=<path/url to your mmdetection config> -P checkpoint=<path/url to your mmdetection model checkpoint>
The resulting segmentation should be written to output.json
and logged as an artifact in the mlflow run.
By default the config and checkpoint paths are cached when specified as a url. Therefore, the CACHE_FOLDER
environment variable must point to an existing folder that can be used for caching the files.
The wrapper is used to deploy any mmdetection methods in the SegServe runtime environment. SegServe can be used to host 3rd party segmentation algorithms and execute them on a central computer while providing a REST interface for clients. Therefore, end-users do not need any powerful hardware/GPU.