wormpose.machine_learning.eval_data_generator

Generates evaluation data: random real processed images with labels and save them to a Tfrecord file

wormpose.machine_learning.eval_data_generator.generate(dataset: wormpose.dataset.loader.Dataset, num_samples: int, theta_dims: int, file_pattern: str) → int

Generates evaluation dataset composed of processed real images, saved to a .TFrecord

Parameters

• dataset – WormPose Dataset

• num_samples – How many images to generate

• theta_dims – Dimensions of theta for the labels

• file_pattern – Path of the output files with “index” variable example: “path_to_out/eval_{index}.tfrecord”

Returns

How many samples where actually generated (if there is less data than the requested num_samples)