Fall detection with image processing and balance sensor data using Arduino Nicla Vision

Help
#1

Hello, how can I do both visual data training and sensor data training? Is training done on only one data type?
I would also be happy if you have any suggestions that could help me with this project.

#2

Hi @abdullahs,

Yes, at this time, only one data type and model is supported per project, and you can only run one impulse per board (or process if you’re running on e.g. Linux). You will likely need to create a custom model (e.g. with TensorFlow or PyTorch) that can handle both types of input. With that, you can still use the Edge Impulse BYOM to convert your model to a C++ or Arduino library.