bobby/open_dbm
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix code refactoring in intensity

Browse Source
This commit is contained in:
jordi.hasianta
2022-09-15 21:34:08 +07:00
parent 4218ea4bef
commit 8550ac7954
1 changed files with 47 additions and 106 deletions
Download Patch File Download Diff File Expand all files Collapse all files

143
opendbm/dbm_lib/dbm_features/raw_features/audio/intensity.py
View File

@@ -1,15 +1,13 @@
"""
file_name: gne
file_name: intensity
project_name: DBM
created: 2020-20-07
"""
import glob
import logging
import os
from os.path import join
import more_itertools as mit
import numpy as np
import pandas as pd
import parselmouth
@@ -19,115 +17,67 @@ from opendbm.dbm_lib.dbm_features.raw_features.util import util as ut
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger()
gne_dir = "acoustic/glottal_noise"
ff_dir = "acoustic/pitch"
csv_ext = "_gne.csv"
intensity_dir = "acoustic/intensity"
csv_ext = "_intensity.csv"
error_txt = "error: length less than 0.064"
def gne_ratio(sound):
def intensity_score(path):
"""
Using parselmouth library fetching glottal noise excitation ratio
Using parselmouth library fetching Intensity
Args:
sound: parselmouth object
path: (.wav) audio file location
Returns:
(list) list of gne ratio for each voice frame
(list) list of Intensity for each voice frame
"""
harmonicity_gne = sound.to_harmonicity_gne()
gne_all_bands = harmonicity_gne.values
gne_all_bands = np.where(gne_all_bands == -200, np.NaN, gne_all_bands)
gne = np.nanmax(
gne_all_bands
) # following http://www.fon.hum.uva.nl/rob/NKI_TEVA/TEVA/HTML/NKI_TEVA.pdf
return gne
sound_pat = parselmouth.Sound(path)
intensity = sound_pat.to_intensity(time_step=0.001)
return intensity.values[0]
def empty_gne(video_uri, out_loc, fl_name, r_config, error_txt, save=True):
def calc_intensity(video_uri, audio_file, out_loc, fl_name, r_config, save=True):
"""
Preparing empty GNE matrix if something fails
"""
cols = ["Frames", r_config.aco_gne, r_config.err_reason]
out_val = [[np.nan, np.nan, error_txt]]
df_gne = pd.DataFrame(out_val, columns=cols)
df_gne["dbm_master_url"] = video_uri
if save:
logger.info("Saving Output file {} ".format(out_loc))
ut.save_output(df_gne, out_loc, fl_name, gne_dir, csv_ext)
return df_gne
def segment_gne(com_speech_sort, voiced_yes, voiced_no, gne_all_frames, audio_file):
"""
calculating gne for each voice segment
"""
snd = parselmouth.Sound(audio_file)
pitch = snd.to_pitch(time_step=0.001)
for idx, vs in enumerate(com_speech_sort):
try:
max_gne = np.NaN
if vs in voiced_yes and len(vs) > 1:
start_time = pitch.get_time_from_frame_number(vs[0])
end_time = pitch.get_time_from_frame_number(vs[-1])
snd_start = int(snd.get_frame_number_from_time(start_time))
snd_end = int(snd.get_frame_number_from_time(end_time))
samples = parselmouth.Sound(snd.as_array()[0][snd_start:snd_end])
max_gne = gne_ratio(samples)
except:
pass
gne_all_frames[idx] = max_gne
return gne_all_frames
def calc_gne(video_uri, audio_file, out_loc, fl_name, r_config, save=True, ff_df=None):
"""
Preparing gne matrix
Preparing Intensity matrix
Args:
audio_file: (.wav) parsed audio file
out_loc: (str) Output directory for csv's
"""
dir_path = os.path.join(out_loc, ff_dir)
if os.path.isdir(dir_path) or ff_df is not None:
if ff_df is not None:
voice_seg = ut.process_segment_pitch(ff_df, r_config)
else:
voice_seg = ut.segment_pitch(dir_path, r_config, ff_df=ff_df)
gne_all_frames = [np.NaN] * len(voice_seg[0])
gne_segment_frames = segment_gne(
voice_seg[0], voice_seg[1], voice_seg[2], gne_all_frames, audio_file
)
intensity_frames = intensity_score(audio_file)
df_intensity = pd.DataFrame(intensity_frames, columns=[r_config.aco_int])
df_gne = pd.DataFrame(gne_segment_frames, columns=[r_config.aco_gne])
df_gne[
df_intensity["Frames"] = df_intensity.index
df_intensity["dbm_master_url"] = video_uri
df_intensity[
r_config.err_reason
] = "Pass" # will replace with threshold in future release
df_gne["Frames"] = df_gne.index
df_gne["dbm_master_url"] = video_uri
if save:
logger.info("Saving Output file {} ".format(out_loc))
ut.save_output(df_intensity, out_loc, fl_name, intensity_dir, csv_ext)
return df_intensity
def empty_intensity(video_uri, out_loc, fl_name, r_config, save=True):
"""
Preparing empty Intensity matrix if something fails
"""
cols = ["Frames", r_config.aco_int, r_config.err_reason]
out_val = [[np.nan, np.nan, error_txt]]
df_int = pd.DataFrame(out_val, columns=cols)
df_int["dbm_master_url"] = video_uri
if save:
logger.info("Processing Output file {} ".format(out_loc))
ut.save_output(df_gne, out_loc, fl_name, gne_dir, csv_ext)
return df_gne
else:
error_txt = "error: pitch freq not available"
return empty_gne(video_uri, out_loc, fl_name, r_config, error_txt, save=save)
logger.info("Saving Output file {} ".format(out_loc))
ut.save_output(df_int, out_loc, fl_name, intensity_dir, csv_ext)
return df_int
def run_gne(video_uri, out_dir, r_config, save=True, ff_df=None):
def run_intensity(video_uri, out_dir, r_config, save=True):
"""
Processing all patient's for fetching glottal noise ratio
---------------
---------------
Processing all patient's for fetching Intensity
-------------------
-------------------
Args:
video_uri: video path; r_config: raw variable config object
out_dir: (str) Output directory for processed output
@@ -146,19 +96,10 @@ def run_gne(video_uri, out_dir, r_config, save=True, ff_df=None):
"Output file {} size is less than 0.064sec".format(audio_file)
)
error_txt = "error: length less than 0.064"
df = empty_gne(
video_uri, out_loc, fl_name, r_config, error_txt, save=save
)
df = empty_intensity(video_uri, out_loc, fl_name, r_config, save=save)
else:
df = calc_gne(
video_uri,
audio_file,
out_loc,
fl_name,
r_config,
save=save,
ff_df=ff_df,
df = calc_intensity(
video_uri, audio_file, out_loc, fl_name, r_config, save=save
)
return df
except Exception as e: