SPRADB0 may   2023 AM62A3 , AM62A3-Q1 , AM62A7 , AM62A7-Q1 , AM67A , AM68A , AM69A

 

  1.   1
  2.   Abstract
  3.   Trademarks
  4. 1Introduction
    1. 1.1 Intended Audience
    2. 1.2 Host Machine Information
  5. 2Creating the Dataset
    1. 2.1 Collecting Images
    2. 2.2 Labelling Images
    3. 2.3 Augmenting the Dataset (Optional)
  6. 3Selecting a Model
  7. 4Training the Model
    1. 4.1 Input Optimization (Optional)
  8. 5Compiling the Model
  9. 6Using the Model
  10. 7Building the End Application
    1. 7.1 Optimizing the Application With TI’s Gstreamer Plugins
    2. 7.2 Using a Raw MIPI-CSI2 Camera
  11. 8Summary
  12. 9References

8 Summary

This document described the step-by-step process of building an Edge AI application for vision tasks on TI AM6xA processors, using the AM62A and a retail-scanner demo to provide an example of the process. Following this development flow can greatly accelerate new designs. This document has detailed the TI Processor-specific steps for training, compiling, and running a model, as well as creating a gstreamer pipeline in Linux that leverages hardware accelerators for image processing, without requiring the user to program those accelerators manually. References are provided to point developers to the tools, code, and further guidance to develop the applications that leverage the complexity of the processor without introducing additional design complexity.