Add minimal quickstart

README.md

@@ -35,6 +35,64 @@ External pull requests are welcome, but must be reviewed before they can be merg
 
 For a longer desciption of the branching model, please refer to our [examples](https://github.com/RI-SE/smirk/blob/main/branching.md).
 
+## Quickstart
+Create a python 3.7 virtual environment with your preferred tool e.g. using conda:
+
+```
+$ conda create -n=smirk-env python=3.7
+$ conda activate smirk-env
+```
+
+Installation in interactive mode currently requires [poetry](https://python-poetry.org/). Once poetry is installed SMRIK can be installed as follows:
+
+```
+$ git clone https://github.com/RI-SE/smirk && cd smirk
+$ poetry install
+```
+
+This will allow you to use the smirk CLI command to reproduce the smirk results or try your own configurations.
+
+Use the `-h` option to see additional help.
+
+```
+$ smirk -h
+
+Usage: smirk [OPTIONS] COMMAND [ARGS]...
+
+  Experimental pedestrian emergency breaking ADAS facilitating research
+  on quality assurance of critical components that rely on machine
+  learning.
+
+Options:
+  -v, --version  Show the version and exit.
+  -h, --help     Show this message and exit.
+
+Commands:
+  detect    Training and evaluation of pedestrian detection models.
+  generate  Data generation
+  safety    Outlier detection training and evaluation.
+  test      System simulator testing.
+```
+
+The `-h` option is available for sub-commands as well e.g.:
+
+```
+$ smirk detect eval -h
+
+Usage: smirk detect eval [OPTIONS]
+
+  Evalute pedestrian detection model.
+
+Options:
+  -d, --data DIRECTORY  Path to dataset directory  [required]
+  -w, --weights FILE    Path to model weights  [required]
+  --conf FLOAT          Confidence threshold  [required]
+  --batch-size INTEGER  Batch size
+  -h, --help            Show this message and exit.
+```
+
+Further instructions on reproducing the results and running custom experiments will be provided in the [docs](/docs/README.md).
+
 ## References
 <a id="1">[1]</a> https://futurium.ec.europa.eu/en/european-ai-alliance/pages/altai-assessment-list-trustworthy-artificial-intelligence)