Important
We are currently writing new tutorials and updating the old ones, meaning that some details may be outdated. We appreciate your understanding!
Important
All our tutorials requires DepthAI v3. If you're looking for tutorials using DepthAI v2, please look here.
This repository contains step by step tutorials on how to:
- Train your own Machine Learning (ML) model - either on default or custom dataset.
- Convert the ML model so it's compatible with the DepthAI v3 platform.
DepthAI is the platform for Spatial AI. We've prepared several demos for various purposes; check out this repository to learn more about them.
depthai-ml-training/
├── conversion/ - Folder with tutorials about ML model conversion
│ ├── ...
│ └── README.md - Describes the conversion tutorials
├── training/ - Folder with tutorials showing how to train a ML model
│ ├── others/ - Folder with tutorials featuring ML model training using other means than LuxonisTrain, such as Ultralytics or TensorFlow
│ │ ├── image-classification/
│ │ ├── object-detection/
│ │ └── semantic-segmentation/
│ ├── ...
│ └── README.md - Describes all the training turorials
├── LICENSE
└── README.md - Describes the whole repository structure and key features
Google Colaboratory allows you to train neural models yourself using their fast GPU instances, and in some cases (depending on the dependencies of the training frameworks), even using the Google TPU - all for free!
It is exactly intended for this proof of concept and initial research. And if you hit the limits of the free account, you can upgrade to a Pro version of Google Colab for only $10/month.
You can browse (and open/use) all the Jupyter Notebooks in this repo by clicking on 
All the tutorials are licensed under the MIT license.
We welcome contributions! Whether it's reporting bugs, improving documentation or tutorials, or adding new tutorials, your help is much appreciated. Please create a pull request (here's how to do it) and assign anyone from the Luxonis team to review the suggested changes. Cheers!