8.1.1.1.1.9. blueoil.cmd.predict

8.1.1.1.1.9.1. Module Contents

8.1.1.1.1.9.1.1. Functions

_get_images(filenames, pre_processor, data_format)

_all_image_files(directory)

_run(input_dir, output_dir, config, restore_path, save_images)

run(input_dir, output_dir, experiment_id, config_file, restore_path, save_images)

predict(input_dir, output_dir, experiment_id, config_file=None, checkpoint=None, save_images=True)

Make predictions from input dir images by using trained model.

blueoil.cmd.predict.DUMMY_FILENAME = DUMMY_FILE
blueoil.cmd.predict._get_images(filenames, pre_processor, data_format)
blueoil.cmd.predict._all_image_files(directory)
blueoil.cmd.predict._run(input_dir, output_dir, config, restore_path, save_images)
blueoil.cmd.predict.run(input_dir, output_dir, experiment_id, config_file, restore_path, save_images)
blueoil.cmd.predict.predict(input_dir, output_dir, experiment_id, config_file=None, checkpoint=None, save_images=True)

Make predictions from input dir images by using trained model. Save the predictions npy, json, images results to output dir. npy: {output_dir}/npy/{batch number}.npy json: {output_dir}/json/{batch number}.json images: {output_dir}/images/{some type}/{input image file name}