Based on official Tensorflow installation tutorial and @nicknochnack youtube tutorial I'm creating tool that will hopefully help you with setting up basic environment to start your object detection project. It creates every necessary directory, creates virtual environment, clones required github repositories, installs object_detection_api and pretrained models.
- In Terminal:
python3 createRequiredPaths.py
- In Terminal:
source venv/bin/activate
- In Terminal:
python createRequiredPaths.py
- In Terminal:
python createLabelMap.py
and follow steps - Put train and test data in respective folders (workspace/image...) ???
- In Terminal:
python trainer.py
- In terminal:
<command generated by trainer.py>
Nicholas Renotte:
https://www.youtube.com/@NicholasRenotte
https://youtu.be/IOI0o3Cxv9Q
https://github.com/nicknochnack
Tensorflow installation tutorial:
https://tensorflow-object-detection-api-tutorial.readthedocs.io/en/latest/install.html
OS: Ununtu 22.04