added low level background noise (79f85aae) · Commits · IVAS Codec Public Collaboration / IVAS Processing Scripts

README.md

+5 −2

Original line number	Diff line number	Diff line
		@@ -239,10 +239,13 @@ input:
		# preamble_noise: true
		### Additive background noise
		# background_noise:
		### REQUIRED: SNR for background noise in dB
		### SNR for background noise in dB; REQUIRED for prerecorded background noise and ignored for low level noise
		# snr: 10
		### REQUIRED: Path to background noise, must have same format and sampling rate as input signal(s)
		### REQUIRED: Either background noise path or low level noise flag
		### Path to background noise, must have same format and sampling rate as input signal(s); default = null
		# background_noise_path: ".../noise.wav"
		### Flag for using low level [-4,+4] background noise; default = false
		# low_level_noise: true
		```

		</details>

examples/TEMPLATE.yml

+5 −2

Original line number	Diff line number	Diff line
		@@ -106,10 +106,13 @@ input:
		# preamble_noise: true
		### Additive background noise
		# background_noise:
		### REQUIRED: SNR for background noise in dB
		### SNR for background noise in dB; REQUIRED for prerecorded background noise and ignored for low level noise
		# snr: 10
		### REQUIRED: Path to background noise, must have same format and sampling rate as input signal(s)
		### REQUIRED: Either background noise path or low level noise flag
		### Path to background noise, must have same format and sampling rate as input signal(s); default = null
		# background_noise_path: ".../noise.wav"
		### Flag for using low level [-4,+4] background noise; default = false
		# low_level_noise: true

		#################################################
		### Bitstream processing

ivas_processing_scripts/audiotools/audioarray.py

+9 −2

Original line number	Diff line number	Diff line
		@@ -52,6 +52,7 @@ def trim(
		limits: Optional[Tuple[int, int]] = None,
		pad_noise: Optional[bool] = False,
		samples: Optional[bool] = False,
		seed: Optional[int] = None,
		) -> np.ndarray:
		"""
		Trim an audio array
		@@ -88,6 +89,9 @@ def trim(
		if pre_trim < 0:
		if pad_noise:
		# pad with uniformly distributed noise between -4 and 4
		if seed:
		np.random.seed(seed)
		else:
		np.random.seed(SEED_PADDING)
		noise = np.random.randint(
		low=-4, high=5, size=(np.abs(pre_trim), np.shape(x)[1])
		@@ -101,6 +105,9 @@ def trim(
		if post_trim < 0:
		if pad_noise:
		# pad with uniformly distributed noise between -4 and 4
		if seed:
		np.random.seed(seed)
		else:
		np.random.seed(SEED_PADDING)
		noise = np.random.randint(
		low=-4, high=5, size=(np.abs(post_trim), np.shape(x)[1])

ivas_processing_scripts/processing/chains.py

+4 −9

Original line number	Diff line number	Diff line
		@@ -40,7 +40,7 @@ from ivas_processing_scripts.processing.ivas import IVAS
		from ivas_processing_scripts.processing.postprocessing import Postprocessing
		from ivas_processing_scripts.processing.preprocessing import Preprocessing
		from ivas_processing_scripts.processing.preprocessing_2 import Preprocessing2
		from ivas_processing_scripts.utils import list_audio
		from ivas_processing_scripts.utils import get_abs_path, list_audio


		def init_processing_chains(cfg: TestConfig) -> None:
		@@ -133,6 +133,8 @@ def get_preprocessing_2(cfg: TestConfig) -> dict:
		}

		pre2_cfg = cfg.preprocessing_2

		# set up background noise
		background_cfg = pre2_cfg.get("background_noise", None)
		if background_cfg:
		background = {
		@@ -140,6 +142,7 @@ def get_preprocessing_2(cfg: TestConfig) -> dict:
		"background_noise_path": get_abs_path(
		background_cfg.get("background_noise_path", None)
		),
		"low_level_noise": background_cfg.get("low_level_noise", False),
		"seed_delay": cfg.prerun_seed,
		"master_seed": cfg.master_seed,
		"output_fmt": cfg.postprocessing["fmt"],
		@@ -414,11 +417,3 @@ def get_processing_chain(
		)

		return chain


		def get_abs_path(rel_path):
		if rel_path is not None:
		abs_path = Path(rel_path).resolve().absolute()
		else:
		abs_path = None
		return abs_path

ivas_processing_scripts/processing/config.py

+33 −6

Original line number	Diff line number	Diff line
		@@ -46,6 +46,7 @@ from ivas_processing_scripts.constants import (
		REQUIRED_KEYS_MNRU,
		SUPPORTED_CONDITIONS,
		)
		from ivas_processing_scripts.utils import get_abs_path


		def merge_dicts(base: dict, other: dict) -> None:
		@@ -139,12 +140,38 @@ class TestConfig:
		raise KeyError(f"The following key(s) must be specified : {MISSING_KEYS}")

		# validate preprocessing on concatenated file stage
		if (pre_proc_2 := getattr(cfg, "preprocessing_2", None)) is not None:
		bg_noise_folder = Path(pre_proc_2["background_noise_path"]).parent
		if (
		bg_noise_folder.resolve().absolute()
		== cfg.input_path.resolve().absolute()
		if (pre_proc_2 := cfg.get("preprocessing_2", None)) is not None and (
		bg_noise := pre_proc_2.get("background_noise", None)
		) is not None:

		# check if low level flag or path is given
		if not bg_noise.get("background_noise_path", None) and not bg_noise.get(
		"low_level_noise", False
		):
		raise ValueError(
		"Path to prerecorded noise or low level noise flag has to be provided for background noise"
		)

		if bg_noise.get("background_noise_path", None) and bg_noise.get(
		"low_level_noise", False
		):
		raise ValueError(
		"Only prerecorded or low level background noise possible, not both"
		)

		# check if path of background noise exists or is in input folder
		if bg_noise.get("background_noise_path", None):
		# check snr
		if not bg_noise.get("snr", None):
		raise ValueError("SNR has to be specified for background noise")

		if not Path(bg_noise["background_noise_path"]).exists():
		raise ValueError(
		f"Background noise path {bg_noise['background_noise_path']} does not exist"
		)

		bg_noise_folder = Path(bg_noise["background_noise_path"]).parent
		if get_abs_path(bg_noise_folder) == get_abs_path(cfg["input_path"]):
		raise ValueError(
		"Background noise file has to be placed outside the input folder!"
		)