auto-annotate

💻 run locally

Remember don't install your dependencies globally, use venv.

# clone repository and navigate to root directory
git clone [email protected]:roboflow-ai/auto-annotate.git
cd auto-annotate

# setup python environment and activate it
python3 -m venv venv
source venv/bin/activate

# install
pip install -e .

🏃 hit the ground running

Warning Your ROBOFLOW_API_KEY is the secret. Do not commit it to your repository, especially if it is public.

export ROBOFLOW_API_KEY= ... 
python -m a2.annotate \
--source_image_directory ... \
--target_annotation_directory ... \
--roboflow_project_id ... \
--roboflow_project_version ...

🪄 label assist

auto-annotate is perfect if you want to automatically label large batches of data fast. However, if you prefare more control over what detections fall into your dataset, you can use Label Assist in our web interface. You will use the same pre-trained model, but have the ability to correct AI suggestions on the fly.