This directory contains scripts and tools for training, evaluating and deploying HAR models using TensorFlow & STM32Cube.AI.
Under training you can find a step by step guide plus the necessary scripts and tools to train and evaluate the HAR models on custom and public datasets.
Under evaluation you can find a step by step guide plus the necessary scripts and tools for evaluating your model performances if datasets are provided.
Under benchmarking you can find a step by step guide plus the necessary scripts and tools to benchmark your model using STM32Cube.AI through our STM32Cube.AI Developer Cloud Services or from the local download.
Under deployment you can find a step by step guide plus the necessary scripts and tools to deploy your own pre-trained HAR model on your STM32 board using STM32Cube.AI.
You can also use a pretrained model from our Human Activity Recognition STM32 model zoo, check out the available models in human_activity_recognition/pretrained_models directory.