Mastering Computer Vision with TensorFlow 2.x

Running an inference test

This step involves exporting trained inference graphs and running inference tests. The inference is done using the following Python command:

!python /content/models/research/object_detection/export_inference_graph.py \ --input_type=image_tensor \ --pipeline_config_path={pipeline_fname} \ --output_directory={output_directory} \ --trained_checkpoint_prefix={last_model_path}

Here, last_model_path is model_dir, where model checkpoints are stored during training, and pipeline_fname is the model path and config file. Checkpoints cover the value of parameters used by the model during training. The following image shows the output of the four different models developed during training. These run one by one by executing the ...