recipes.icassp_2023.baseline.evaluate module

Evaluate the enhanced signals using a combination of HASPI and HASQI metrics

class recipes.icassp_2023.baseline.evaluate.ResultsFile(file_name)

Bases: object

Class to write results to a CSV file

add_result(scene: str, listener: str, score: float, haspi: float, hasqi: float)

Add a result to the CSV file

write_header()

recipes.icassp_2023.baseline.evaluate.amplify_signal(signal, audiogram: Audiogram, enhancer, compressor)

Amplify signal for a given audiogram

recipes.icassp_2023.baseline.evaluate.compute_metric(metric, signal: ndarray, ref: ndarray, listener: Listener, sample_rate: float)

Compute HASPI or HASQI metric

recipes.icassp_2023.baseline.evaluate.make_scene_listener_list(scenes_listeners, small_test=False)

Make the list of scene-listener pairing to process

recipes.icassp_2023.baseline.evaluate.run_calculate_si(cfg: DictConfig) → None

Evaluate the enhanced signals using a combination of HASPI and HASQI metrics

recipes.icassp_2023.baseline.evaluate.set_scene_seed(scene)

Set a seed that is unique for the given scene