K210 learning notes (IV) k210 runs multiple models at the same time

Preface

In the use of K210 In the process , It is inevitable to run multiple models , Rather than running at the same time , Nothing runs at the same time , It's just time-sharing , If the memory is enough or the model is small , Then we can burn directly in the model area , The starting address is usually 0x300000, Larger models can be placed in SD card . Refer to the post ：K210 MaixPy From entry to flight –AI Visual – Complete tutorial （ And some small problems, such as insufficient memory ）

One 、 Running multiple models at the same time

1. There are two ways , If you have enough memory , Just load a few models into memory at a time , Then the reasoning is run separately in time .
2. If there is not enough memory , After loading the first model , Cancellation , Free memory , Then run the second model after loading .

Two 、 routine

For example, this face recognition routine uses three models at once , Face detection , Feature point recognition, etc , Look at what others wrote , Just change it according to your own needs .
Code address ：https://github.com/sipeed/MaixPy_scripts/blob/master/machine_vision/face_recognization/demo_face_recognization.py

Refer to the post ：
MaixPy Document Introduction
K210 MaixPy From entry to flight –AI Visual – Complete tutorial （ And some small problems, such as insufficient memory ）