Mastering Computer Vision with TensorFlow 2.x

Converting a TensorFlow model into tflite format

This section will describe how to convert a TensorFlow model into a tflite format. If we don't do this, the model we've developed can be used for inference in your local PC, but can't be deployed to edge devices for real-time inference. Three methods will be described regarding this conversion:

• The Python API, which is used in local PCs for tflite conversion
• Google Colab using tflite conversion
• Google Colab using toco

Since this is an object detection conversion, where our model was developed based on the TensorFlow object detection API, we will use the toco method in Google Colab.