Deploying tensorflowlite object detection model

[modulo3see]

Hi, I am using STM32F746G Discovery Board to deploy an object detection model to detect person in an image. The model size of the object detection model is less than 1Mb. The implementation would be like to make an inference in the microcontroller like access the jpg images in the sd card (since i dont have a camera module) and display it in the LCD together with its prediction like bounding boxes on the image in case there is a person in it.

 

I used the STM32Cube MX and enabled XCUBE AI then upload the pre trained model which is the ssd mobile net v2 but I am having trouble making the configuration for displaying image, accessing files in sd card, decoding the jpg images from the sd card in STM32Cube MX land I also don't know how to use the XCUBE AI application template like the methods ai_run(), acquire_and_process_data(ai_i8* data[]) and others. 

 

[Julian E.]

Hello @modulo3see ,

 

I think you could start by using the ST Edge AI Developer Cloud to have an example of project that use the CubeAI Function. In the last step of the dev cloud, you can select your target and generate a STM32CubeIDE project. By default this project use a random input for your model and send by serial some information like the inference time. You could check the code to understand more how to use cube AI.

dev could: https://stedgeai-dc.st.com/home

documentation: https://stedgeai-dc.st.com/documentation

dev cloud getting started: https://wiki.st.com/stm32mcu/wiki/AI:Getting_started_with_STM32Cube.AI_Developer_Cloud 

 

Once you understand how to use a model with random input, I think you can start to look to solve your other issues related to display or accessing a sd card. I think it will be better to open seprate question in the related wikis.

 

There is also the documentation of CubeAI that you can find like explain here: https://wiki.st.com/stm32mcu/wiki/AI:X-CUBE-AI_documentation#X-CUBE-AI_embedded_documentation_access 

 

We have a function pack FP-AI-VISION1 which is quite similar to what you are trying to do. You may find useful information in the user manual: https://www.st.com/en/embedded-software/fp-ai-vision1.html#documentation

 

Have a good day,

Julian