5462629459

pending completion

Build # 5462629459

Build Type

Pull #899

github

Committed by

web-flow

Commit Message

Merge e759dacb6 into c94f0e331

Pull Request Pull Request #899: [DONE] Ptitloup/feature new encoding

Run Details

400 of 400 new or added lines in 14 files covered. (100.0%)

8968 of 12675 relevant lines covered (70.75%)

0.71 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

0.0

/pod/video_encode_transcript/transcript_model.py

import numpy as np
import shlex
import subprocess
import json

import sys
import os
from timeit import default_timer as timer
import datetime as dt
from datetime import timedelta

from webvtt import WebVTT, Caption

try:
    from shhlex import quote
except ImportError:
    from pipes import quote

import logging

from ..custom import settings_local

DEBUG = getattr(settings_local, "DEBUG", False)

TRANSCRIPTION_MODEL_PARAM = getattr(settings_local, "TRANSCRIPTION_MODEL_PARAM", False)
USE_TRANSCRIPTION = getattr(settings_local, "USE_TRANSCRIPTION", False)
if USE_TRANSCRIPTION:
    TRANSCRIPTION_TYPE = getattr(settings_local, "TRANSCRIPTION_TYPE", "VOSK")
    if TRANSCRIPTION_TYPE == "VOSK":
        from vosk import Model, KaldiRecognizer
    elif TRANSCRIPTION_TYPE == "STT":
        from stt import Model

TRANSCRIPTION_NORMALIZE = getattr(settings_local, "TRANSCRIPTION_NORMALIZE", False)
TRANSCRIPTION_NORMALIZE_TARGET_LEVEL = getattr(
    settings_local, "TRANSCRIPTION_NORMALIZE_TARGET_LEVEL", -16.0
)

TRANSCRIPTION_AUDIO_SPLIT_TIME = getattr(
    settings_local, "TRANSCRIPTION_AUDIO_SPLIT_TIME", 600
)  # 10min
# time in sec for phrase length
TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH = getattr(
    settings_local, "TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH", 3
)
TRANSCRIPTION_STT_SENTENCE_BLANK_SPLIT_TIME = getattr(
    settings_local, "TRANSCRIPTION_STT_SENTENCE_BLANK_SPLIT_TIME", 0.5
)
log = logging.getLogger(__name__)


def get_model(lang):
    """Get model for STT or Vosk software to transcript audio."""
    transript_model = Model(TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["model"])
    if TRANSCRIPTION_TYPE == "STT":
        if TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang].get("beam_width"):
            transript_model.setBeamWidth(
                TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["beam_width"]
            )
        if TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang].get("scorer"):
            print(
                "Loading scorer from files {}".format(
                    TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["scorer"]
                ),
                file=sys.stderr,
            )
            scorer_load_start = timer()
            transript_model.enableExternalScorer(
                TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["scorer"]
            )
            scorer_load_end = timer() - scorer_load_start
            print("Loaded scorer in {:.3}s.".format(scorer_load_end), file=sys.stderr)
            if TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang].get(
                "lm_alpha"
            ) and TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang].get("lm_beta"):
                transript_model.setScorerAlphaBeta(
                    TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["lm_alpha"],
                    TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["lm_beta"],
                )
    return transript_model


def start_transcripting(mp3filepath, duration, lang):
    """
    Normalize the audio if set, get the model according to the lang and start transcript.
    """
    if TRANSCRIPTION_NORMALIZE:
        mp3filepath = normalize_mp3(mp3filepath)
    transript_model = get_model(lang)
    msg, webvtt, all_text = start_main_transcript(
        mp3filepath, duration, transript_model
    )
    if DEBUG:
        print(msg)
        print(webvtt)
        print("\n%s\n" % all_text)

    return msg, webvtt


def start_main_transcript(mp3filepath, duration, transript_model):
    """Call transcription depending software type."""
    if TRANSCRIPTION_TYPE == "STT":
        msg, webvtt, all_text = main_stt_transcript(
            mp3filepath, duration, transript_model
        )
    elif TRANSCRIPTION_TYPE == "VOSK":
        msg, webvtt, all_text = main_vosk_transcript(
            mp3filepath, duration, transript_model
        )
    return msg, webvtt, all_text


def convert_samplerate(audio_path, desired_sample_rate, trim_start, duration):
    """Convert audio to subaudio and add good sample rate."""
    sox_cmd = "sox {} --type raw --bits 16 --channels 1 --rate {} ".format(
        quote(audio_path), desired_sample_rate
    )
    sox_cmd += "--encoding signed-integer --endian little --compression 0.0 "
    sox_cmd += "--no-dither - trim {} {}".format(trim_start, duration)

    try:
        output = subprocess.check_output(shlex.split(sox_cmd), stderr=subprocess.PIPE)

    except subprocess.CalledProcessError as e:
        raise RuntimeError("SoX returned non-zero status: {}".format(e.stderr))
    except OSError as e:
        raise OSError(
            e.errno,
            "SoX not found, use {}hz files or install it: {}".format(
                desired_sample_rate, e.strerror
            ),
        )

    return np.frombuffer(output, np.int16)


def normalize_mp3(mp3filepath):
    """Normalize the audio to good format and sound level."""
    filename, file_extension = os.path.splitext(mp3filepath)
    mp3normfile = "{}{}{}".format(filename, "_norm", file_extension)
    normalize_cmd = "ffmpeg-normalize {} ".format(quote(mp3filepath))
    normalize_cmd += "-c:a libmp3lame -b:a 192k --normalization-type ebu "
    # normalize_cmd += \
    # '--loudness-range-target 7.0 --true-peak 0.0 --offset 0.0 '
    normalize_cmd += "--target-level {} -f -o {}".format(
        TRANSCRIPTION_NORMALIZE_TARGET_LEVEL, quote(mp3normfile)
    )
    if DEBUG:
        print(normalize_cmd)
    try:
        subprocess.check_output(shlex.split(normalize_cmd), stderr=subprocess.PIPE)
        return mp3normfile
    except subprocess.CalledProcessError as e:
        log.error("ffmpeg-normalize returned non-zero status: {}".format(e.stderr))
        return mp3filepath
    except OSError as e:
        log.error("ffmpeg-normalize not found {}".format(e.strerror))
        return mp3filepath


# #################################
# TRANSCRIPT VIDEO : MAIN FUNCTION
# #################################


def convert_vosk_samplerate(audio_path, desired_sample_rate, trim_start, duration):
    """Convert audio to the good sample rate."""
    sox_cmd = "sox {} --type raw --bits 16 --channels 1 --rate {} ".format(
        quote(audio_path), desired_sample_rate
    )
    sox_cmd += "--encoding signed-integer --endian little --compression 0.0 "
    sox_cmd += "--no-dither - trim {} {}".format(trim_start, duration)

    try:
        output = subprocess.Popen(shlex.split(sox_cmd), stdout=subprocess.PIPE)

    except subprocess.CalledProcessError as e:
        raise RuntimeError("SoX returned non-zero status: {}".format(e.stderr))
    except OSError as e:
        raise OSError(
            e.errno,
            "SoX not found, use {}hz files or install it: {}".format(
                desired_sample_rate, e.strerror
            ),
        )
    return output


def get_word_result_from_data(results, audio, rec):
    """Get subsound from audio and add transcription to result parameter."""
    while True:
        data = audio.stdout.read(4000)
        if len(data) == 0:
            break
        if rec.AcceptWaveform(data):
            results.append(rec.Result())
    results.append(rec.Result())


def words_to_vtt(
    words,
    start_trim,
    duration,
    is_first_caption,
    text_caption,
    start_caption,
    last_word_added,
    all_text,
    webvtt,
):
    """Convert word and time to webvtt captions."""
    for index, word in enumerate(words):
        start_key = "start_time"
        word_duration = word.get("duration", 0)
        last_word = words[-1]
        last_word_duration = last_word.get("duration", 0)
        if TRANSCRIPTION_TYPE == "VOSK":
            start_key = "start"
            word_duration = word["end"] - word["start"]
            last_word_duration = words[-1]["end"] - words[-1]["start"]
        next_word = None
        blank_duration = 0
        if word != words[-1] and (index + 1) < len(words):
            next_word = words[index + 1]
            blank_duration = ((next_word[start_key]) - start_caption) - (
                ((word[start_key]) - start_caption) + word_duration
            )
        all_text += word["word"] + " "
        # word : <class 'dict'> {'word': 'bonjour', 'start ':
        # 0.58, 'duration': 7.34}
        text_caption.append(word["word"])
        if not (
            (((word[start_key]) - start_caption) < TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH)
            and (
                next_word is not None
                and (blank_duration < TRANSCRIPTION_STT_SENTENCE_BLANK_SPLIT_TIME)
            )
        ):
            # on créé le caption
            if is_first_caption:
                # A revoir, fusion de la nouvelle ligne avec
                # l'ancienne...
                is_first_caption = False
                text_caption = get_text_caption(text_caption, last_word_added)

            stop_caption = word[start_key] + word_duration

            # on evite le chevauchement
            change_previous_end_caption(webvtt, start_caption)

            caption = Caption(
                format_time_caption(start_caption),
                format_time_caption(stop_caption),
                " ".join(text_caption),
            )

            webvtt.captions.append(caption)
            # on remet tout à zero pour la prochaine phrase
            start_caption = word[start_key]
            text_caption = []
            last_word_added = word["word"]
    if start_trim + TRANSCRIPTION_AUDIO_SPLIT_TIME > duration:
        # on ajoute ici la dernière phrase de la vidéo
        stop_caption = start_trim + words[-1][start_key] + last_word_duration
        caption = Caption(
            format_time_caption(start_caption),
            format_time_caption(stop_caption),
            " ".join(text_caption),
        )
        webvtt.captions.append(caption)
    return all_text, webvtt


def main_vosk_transcript(norm_mp3_file, duration, transript_model):
    """Vosk transcription."""
    msg = ""
    inference_start = timer()
    msg += "\nInference start %0.3fs." % inference_start
    desired_sample_rate = 16000

    rec = KaldiRecognizer(transript_model, desired_sample_rate)
    rec.SetWords(True)

    webvtt = WebVTT()
    all_text = ""
    for start_trim in range(0, duration, TRANSCRIPTION_AUDIO_SPLIT_TIME):
        audio = convert_vosk_samplerate(
            norm_mp3_file,
            desired_sample_rate,
            start_trim,
            TRANSCRIPTION_AUDIO_SPLIT_TIME,  # dur
        )
        msg += "\nRunning inference."
        results = []
        get_word_result_from_data(results, audio, rec)
        for res in results:
            words = json.loads(res).get("result")
            text = json.loads(res).get("text")
            if not words:
                continue
            start_caption = words[0]["start"]
            stop_caption = words[-1]["end"]
            caption = Caption(
                format_time_caption(start_caption),
                format_time_caption(stop_caption),
                text,
            )
            webvtt.captions.append(caption)
            """
            text_caption = []
            is_first_caption = True
            all_text, webvtt = words_to_vtt(
                words,
                start_trim,
                duration,
                is_first_caption,
                text_caption,
                start_caption,
                last_word_added,
                all_text,
                webvtt,
            )
            """
    inference_end = timer() - inference_start

    msg += "\nInference took %0.3fs." % inference_end
    return msg, webvtt, all_text


def main_stt_transcript(norm_mp3_file, duration, transript_model):
    """STT transcription."""
    msg = ""
    inference_start = timer()
    msg += "\nInference start %0.3fs." % inference_start
    desired_sample_rate = transript_model.sampleRate()
    webvtt = WebVTT()
    last_word_added = ""
    metadata = None
    all_text = ""
    for start_trim in range(0, duration, TRANSCRIPTION_AUDIO_SPLIT_TIME):
        end_trim = (
            duration
            if start_trim + TRANSCRIPTION_AUDIO_SPLIT_TIME > duration
            else (
                start_trim
                + TRANSCRIPTION_AUDIO_SPLIT_TIME
                + TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH
            )
        )

        dur = (
            (TRANSCRIPTION_AUDIO_SPLIT_TIME + TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH)
            if (
                (
                    start_trim
                    + TRANSCRIPTION_AUDIO_SPLIT_TIME
                    + TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH
                )
                < duration
            )
            else (duration - start_trim)
        )

        msg += "\ntake audio from %s to %s - %s" % (start_trim, end_trim, dur)

        audio = convert_samplerate(norm_mp3_file, desired_sample_rate, start_trim, dur)
        msg += "\nRunning inference."

        metadata = transript_model.sttWithMetadata(audio)

        for transcript in metadata.transcripts:
            msg += "\nConfidence : %s" % transcript.confidence
            words = words_from_candidate_transcript(transcript)
            start_caption = start_trim + words[0]["start_time"]
            text_caption = []
            is_first_caption = True
            all_text, webvtt = words_to_vtt(
                words,
                start_trim,
                duration,
                is_first_caption,
                text_caption,
                start_caption,
                last_word_added,
                all_text,
                webvtt,
            )
    inference_end = timer() - inference_start

    msg += "\nInference took %0.3fs." % inference_end
    return msg, webvtt, all_text


def change_previous_end_caption(webvtt, start_caption):
    """Change the end time for caption."""
    if len(webvtt.captions) > 0:
        prev_end = dt.datetime.strptime(webvtt.captions[-1].end, "%H:%M:%S.%f")
        td_prev_end = timedelta(
            hours=prev_end.hour,
            minutes=prev_end.minute,
            seconds=prev_end.second,
            microseconds=prev_end.microsecond,
        ).total_seconds()
        if td_prev_end > start_caption:
            webvtt.captions[-1].end = format_time_caption(start_caption)


def format_time_caption(time_caption):
    """Format time for webvtt caption."""
    return (
        dt.datetime.utcfromtimestamp(0) + timedelta(seconds=float(time_caption))
    ).strftime("%H:%M:%S.%f")[:-3]


def get_text_caption(text_caption, last_word_added):
    """get the text for a caption."""
    try:
        first_index = text_caption.index(last_word_added)
        return text_caption[first_index + 1 :]
    except ValueError:
        return text_caption


def words_from_candidate_transcript(metadata):
    """Get words list from transcription."""
    word = ""
    word_list = []
    word_start_time = 0
    # Loop through each character
    for i, token in enumerate(metadata.tokens):
        # Append character to word if it's not a space
        if token.text != " ":
            if len(word) == 0:
                # Log the start time of the new word
                word_start_time = token.start_time

            word = word + token.text
        # Word boundary is either a space or the last character in the array
        if token.text == " " or i == len(metadata.tokens) - 1:
            word_duration = token.start_time - word_start_time

            if word_duration < 0:
                word_duration = 0

            each_word = dict()
            each_word["word"] = word
            each_word["start_time"] = round(word_start_time, 4)
            each_word["duration"] = round(word_duration, 4)

            word_list.append(each_word)
            # Reset
            word = ""
            word_start_time = 0

    return word_list

1	import numpy as np	×
2	import shlex	×
3	import subprocess	×
4	import json	×
5
6	import sys	×
7	import os	×
8	from timeit import default_timer as timer	×
9	import datetime as dt	×
10	from datetime import timedelta	×
11
12	from webvtt import WebVTT, Caption	×
13
14	try:	×
15	from shhlex import quote	×
16	except ImportError:	×
17	from pipes import quote	×
18
19	import logging	×
20
21	from ..custom import settings_local	×
22
23	DEBUG = getattr(settings_local, "DEBUG", False)	×
24
25	TRANSCRIPTION_MODEL_PARAM = getattr(settings_local, "TRANSCRIPTION_MODEL_PARAM", False)	×
26	USE_TRANSCRIPTION = getattr(settings_local, "USE_TRANSCRIPTION", False)	×
27	if USE_TRANSCRIPTION:	×
28	TRANSCRIPTION_TYPE = getattr(settings_local, "TRANSCRIPTION_TYPE", "VOSK")	×
29	if TRANSCRIPTION_TYPE == "VOSK":	×
30	from vosk import Model, KaldiRecognizer	×
31	elif TRANSCRIPTION_TYPE == "STT":	×
32	from stt import Model	×
33
34	TRANSCRIPTION_NORMALIZE = getattr(settings_local, "TRANSCRIPTION_NORMALIZE", False)	×
35	TRANSCRIPTION_NORMALIZE_TARGET_LEVEL = getattr(	×
36	settings_local, "TRANSCRIPTION_NORMALIZE_TARGET_LEVEL", -16.0
37	)
38
39	TRANSCRIPTION_AUDIO_SPLIT_TIME = getattr(	×
40	settings_local, "TRANSCRIPTION_AUDIO_SPLIT_TIME", 600
41	) # 10min
42	# time in sec for phrase length
43	TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH = getattr(	×
44	settings_local, "TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH", 3
45	)
46	TRANSCRIPTION_STT_SENTENCE_BLANK_SPLIT_TIME = getattr(	×
47	settings_local, "TRANSCRIPTION_STT_SENTENCE_BLANK_SPLIT_TIME", 0.5
48	)
49	log = logging.getLogger(__name__)	×
50
51
52	def get_model(lang):	×
53	"""Get model for STT or Vosk software to transcript audio."""
54	transript_model = Model(TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["model"])	×
55	if TRANSCRIPTION_TYPE == "STT":	×
56	if TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang].get("beam_width"):	×
57	transript_model.setBeamWidth(	×
58	TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["beam_width"]
59	)
60	if TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang].get("scorer"):	×
61	print(	×
62	"Loading scorer from files {}".format(
63	TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["scorer"]
64	),
65	file=sys.stderr,
66	)
67	scorer_load_start = timer()	×
68	transript_model.enableExternalScorer(	×
69	TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["scorer"]
70	)
71	scorer_load_end = timer() - scorer_load_start	×
72	print("Loaded scorer in {:.3}s.".format(scorer_load_end), file=sys.stderr)	×
73	if TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang].get(	×
74	"lm_alpha"
75	) and TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang].get("lm_beta"):
76	transript_model.setScorerAlphaBeta(	×
77	TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["lm_alpha"],
78	TRANSCRIPTION_MODEL_PARAM[TRANSCRIPTION_TYPE][lang]["lm_beta"],
79	)
80	return transript_model	×
81
82
83	def start_transcripting(mp3filepath, duration, lang):	×
84	"""
85	Normalize the audio if set, get the model according to the lang and start transcript.
86	"""
87	if TRANSCRIPTION_NORMALIZE:	×
88	mp3filepath = normalize_mp3(mp3filepath)	×
89	transript_model = get_model(lang)	×
90	msg, webvtt, all_text = start_main_transcript(	×
91	mp3filepath, duration, transript_model
92	)
93	if DEBUG:	×
94	print(msg)	×
95	print(webvtt)	×
96	print("\n%s\n" % all_text)	×
97
98	return msg, webvtt	×
99
100
101	def start_main_transcript(mp3filepath, duration, transript_model):	×
102	"""Call transcription depending software type."""
103	if TRANSCRIPTION_TYPE == "STT":	×
104	msg, webvtt, all_text = main_stt_transcript(	×
105	mp3filepath, duration, transript_model
106	)
107	elif TRANSCRIPTION_TYPE == "VOSK":	×
108	msg, webvtt, all_text = main_vosk_transcript(	×
109	mp3filepath, duration, transript_model
110	)
111	return msg, webvtt, all_text	×
112
113
114	def convert_samplerate(audio_path, desired_sample_rate, trim_start, duration):	×
115	"""Convert audio to subaudio and add good sample rate."""
116	sox_cmd = "sox {} --type raw --bits 16 --channels 1 --rate {} ".format(	×
117	quote(audio_path), desired_sample_rate
118	)
119	sox_cmd += "--encoding signed-integer --endian little --compression 0.0 "	×
120	sox_cmd += "--no-dither - trim {} {}".format(trim_start, duration)	×
121
122	try:	×
123	output = subprocess.check_output(shlex.split(sox_cmd), stderr=subprocess.PIPE)	×
124
125	except subprocess.CalledProcessError as e:	×
126	raise RuntimeError("SoX returned non-zero status: {}".format(e.stderr))	×
127	except OSError as e:	×
128	raise OSError(	×
129	e.errno,
130	"SoX not found, use {}hz files or install it: {}".format(
131	desired_sample_rate, e.strerror
132	),
133	)
134
135	return np.frombuffer(output, np.int16)	×
136
137
138	def normalize_mp3(mp3filepath):	×
139	"""Normalize the audio to good format and sound level."""
140	filename, file_extension = os.path.splitext(mp3filepath)	×
141	mp3normfile = "{}{}{}".format(filename, "_norm", file_extension)	×
142	normalize_cmd = "ffmpeg-normalize {} ".format(quote(mp3filepath))	×
143	normalize_cmd += "-c:a libmp3lame -b:a 192k --normalization-type ebu "	×
144	# normalize_cmd += \
145	# '--loudness-range-target 7.0 --true-peak 0.0 --offset 0.0 '
146	normalize_cmd += "--target-level {} -f -o {}".format(	×
147	TRANSCRIPTION_NORMALIZE_TARGET_LEVEL, quote(mp3normfile)
148	)
149	if DEBUG:	×
150	print(normalize_cmd)	×
151	try:	×
152	subprocess.check_output(shlex.split(normalize_cmd), stderr=subprocess.PIPE)	×
153	return mp3normfile	×
154	except subprocess.CalledProcessError as e:	×
155	log.error("ffmpeg-normalize returned non-zero status: {}".format(e.stderr))	×
156	return mp3filepath	×
157	except OSError as e:	×
158	log.error("ffmpeg-normalize not found {}".format(e.strerror))	×
159	return mp3filepath	×
160
161
162	# #################################
163	# TRANSCRIPT VIDEO : MAIN FUNCTION
164	# #################################
165
166
167	def convert_vosk_samplerate(audio_path, desired_sample_rate, trim_start, duration):	×
168	"""Convert audio to the good sample rate."""
169	sox_cmd = "sox {} --type raw --bits 16 --channels 1 --rate {} ".format(	×
170	quote(audio_path), desired_sample_rate
171	)
172	sox_cmd += "--encoding signed-integer --endian little --compression 0.0 "	×
173	sox_cmd += "--no-dither - trim {} {}".format(trim_start, duration)	×
174
175	try:	×
176	output = subprocess.Popen(shlex.split(sox_cmd), stdout=subprocess.PIPE)	×
177
178	except subprocess.CalledProcessError as e:	×
179	raise RuntimeError("SoX returned non-zero status: {}".format(e.stderr))	×
180	except OSError as e:	×
181	raise OSError(	×
182	e.errno,
183	"SoX not found, use {}hz files or install it: {}".format(
184	desired_sample_rate, e.strerror
185	),
186	)
187	return output	×
188
189
190	def get_word_result_from_data(results, audio, rec):	×
191	"""Get subsound from audio and add transcription to result parameter."""
192	while True:
193	data = audio.stdout.read(4000)	×
194	if len(data) == 0:	×
195	break	×
196	if rec.AcceptWaveform(data):	×
197	results.append(rec.Result())	×
198	results.append(rec.Result())	×
199
200
201	def words_to_vtt(	×
202	words,
203	start_trim,
204	duration,
205	is_first_caption,
206	text_caption,
207	start_caption,
208	last_word_added,
209	all_text,
210	webvtt,
211	):
212	"""Convert word and time to webvtt captions."""
213	for index, word in enumerate(words):	×
214	start_key = "start_time"	×
215	word_duration = word.get("duration", 0)	×
216	last_word = words[-1]	×
217	last_word_duration = last_word.get("duration", 0)	×
218	if TRANSCRIPTION_TYPE == "VOSK":	×
219	start_key = "start"	×
220	word_duration = word["end"] - word["start"]	×
221	last_word_duration = words[-1]["end"] - words[-1]["start"]	×
222	next_word = None	×
223	blank_duration = 0	×
224	if word != words[-1] and (index + 1) < len(words):	×
225	next_word = words[index + 1]	×
226	blank_duration = ((next_word[start_key]) - start_caption) - (	×
227	((word[start_key]) - start_caption) + word_duration
228	)
229	all_text += word["word"] + " "	×
230	# word : <class 'dict'> {'word': 'bonjour', 'start ':
231	# 0.58, 'duration': 7.34}
232	text_caption.append(word["word"])	×
233	if not (	×
234	(((word[start_key]) - start_caption) < TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH)
235	and (
236	next_word is not None
237	and (blank_duration < TRANSCRIPTION_STT_SENTENCE_BLANK_SPLIT_TIME)
238	)
239	):
240	# on créé le caption
241	if is_first_caption:	×
242	# A revoir, fusion de la nouvelle ligne avec
243	# l'ancienne...
244	is_first_caption = False	×
245	text_caption = get_text_caption(text_caption, last_word_added)	×
246
247	stop_caption = word[start_key] + word_duration	×
248
249	# on evite le chevauchement
250	change_previous_end_caption(webvtt, start_caption)	×
251
252	caption = Caption(	×
253	format_time_caption(start_caption),
254	format_time_caption(stop_caption),
255	" ".join(text_caption),
256	)
257
258	webvtt.captions.append(caption)	×
259	# on remet tout à zero pour la prochaine phrase
260	start_caption = word[start_key]	×
261	text_caption = []	×
262	last_word_added = word["word"]	×
263	if start_trim + TRANSCRIPTION_AUDIO_SPLIT_TIME > duration:	×
264	# on ajoute ici la dernière phrase de la vidéo
265	stop_caption = start_trim + words[-1][start_key] + last_word_duration	×
266	caption = Caption(	×
267	format_time_caption(start_caption),
268	format_time_caption(stop_caption),
269	" ".join(text_caption),
270	)
271	webvtt.captions.append(caption)	×
272	return all_text, webvtt	×
273
274
275	def main_vosk_transcript(norm_mp3_file, duration, transript_model):	×
276	"""Vosk transcription."""
277	msg = ""	×
278	inference_start = timer()	×
279	msg += "\nInference start %0.3fs." % inference_start	×
280	desired_sample_rate = 16000	×
281
282	rec = KaldiRecognizer(transript_model, desired_sample_rate)	×
283	rec.SetWords(True)	×
284
285	webvtt = WebVTT()	×
286	all_text = ""	×
287	for start_trim in range(0, duration, TRANSCRIPTION_AUDIO_SPLIT_TIME):	×
288	audio = convert_vosk_samplerate(	×
289	norm_mp3_file,
290	desired_sample_rate,
291	start_trim,
292	TRANSCRIPTION_AUDIO_SPLIT_TIME, # dur
293	)
294	msg += "\nRunning inference."	×
295	results = []	×
296	get_word_result_from_data(results, audio, rec)	×
297	for res in results:	×
298	words = json.loads(res).get("result")	×
299	text = json.loads(res).get("text")	×
300	if not words:	×
301	continue	×
302	start_caption = words[0]["start"]	×
303	stop_caption = words[-1]["end"]	×
304	caption = Caption(	×
305	format_time_caption(start_caption),
306	format_time_caption(stop_caption),
307	text,
308	)
309	webvtt.captions.append(caption)	×
310	"""	×
311	text_caption = []
312	is_first_caption = True
313	all_text, webvtt = words_to_vtt(
314	words,
315	start_trim,
316	duration,
317	is_first_caption,
318	text_caption,
319	start_caption,
320	last_word_added,
321	all_text,
322	webvtt,
323	)
324	"""
325	inference_end = timer() - inference_start	×
326
327	msg += "\nInference took %0.3fs." % inference_end	×
328	return msg, webvtt, all_text	×
329
330
331	def main_stt_transcript(norm_mp3_file, duration, transript_model):	×
332	"""STT transcription."""
333	msg = ""	×
334	inference_start = timer()	×
335	msg += "\nInference start %0.3fs." % inference_start	×
336	desired_sample_rate = transript_model.sampleRate()	×
337	webvtt = WebVTT()	×
338	last_word_added = ""	×
339	metadata = None	×
340	all_text = ""	×
341	for start_trim in range(0, duration, TRANSCRIPTION_AUDIO_SPLIT_TIME):	×
342	end_trim = (	×
343	duration
344	if start_trim + TRANSCRIPTION_AUDIO_SPLIT_TIME > duration
345	else (
346	start_trim
347	+ TRANSCRIPTION_AUDIO_SPLIT_TIME
348	+ TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH
349	)
350	)
351
352	dur = (	×
353	(TRANSCRIPTION_AUDIO_SPLIT_TIME + TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH)
354	if (
355	(
356	start_trim
357	+ TRANSCRIPTION_AUDIO_SPLIT_TIME
358	+ TRANSCRIPTION_STT_SENTENCE_MAX_LENGTH
359	)
360	< duration
361	)
362	else (duration - start_trim)
363	)
364
365	msg += "\ntake audio from %s to %s - %s" % (start_trim, end_trim, dur)	×
366
367	audio = convert_samplerate(norm_mp3_file, desired_sample_rate, start_trim, dur)	×
368	msg += "\nRunning inference."	×
369
370	metadata = transript_model.sttWithMetadata(audio)	×
371
372	for transcript in metadata.transcripts:	×
373	msg += "\nConfidence : %s" % transcript.confidence	×
374	words = words_from_candidate_transcript(transcript)	×
375	start_caption = start_trim + words[0]["start_time"]	×
376	text_caption = []	×
377	is_first_caption = True	×
378	all_text, webvtt = words_to_vtt(	×
379	words,
380	start_trim,
381	duration,
382	is_first_caption,
383	text_caption,
384	start_caption,
385	last_word_added,
386	all_text,
387	webvtt,
388	)
389	inference_end = timer() - inference_start	×
390
391	msg += "\nInference took %0.3fs." % inference_end	×
392	return msg, webvtt, all_text	×
393
394
395	def change_previous_end_caption(webvtt, start_caption):	×
396	"""Change the end time for caption."""
397	if len(webvtt.captions) > 0:	×
398	prev_end = dt.datetime.strptime(webvtt.captions[-1].end, "%H:%M:%S.%f")	×
399	td_prev_end = timedelta(	×
400	hours=prev_end.hour,
401	minutes=prev_end.minute,
402	seconds=prev_end.second,
403	microseconds=prev_end.microsecond,
404	).total_seconds()
405	if td_prev_end > start_caption:	×
406	webvtt.captions[-1].end = format_time_caption(start_caption)	×
407
408
409	def format_time_caption(time_caption):	×
410	"""Format time for webvtt caption."""
411	return (	×
412	dt.datetime.utcfromtimestamp(0) + timedelta(seconds=float(time_caption))
413	).strftime("%H:%M:%S.%f")[:-3]
414
415
416	def get_text_caption(text_caption, last_word_added):	×
417	"""get the text for a caption."""
418	try:	×
419	first_index = text_caption.index(last_word_added)	×
420	return text_caption[first_index + 1 :]	×
421	except ValueError:	×
422	return text_caption	×
423
424
425	def words_from_candidate_transcript(metadata):	×
426	"""Get words list from transcription."""
427	word = ""	×
428	word_list = []	×
429	word_start_time = 0	×
430	# Loop through each character
431	for i, token in enumerate(metadata.tokens):	×
432	# Append character to word if it's not a space
433	if token.text != " ":	×
434	if len(word) == 0:	×
435	# Log the start time of the new word
436	word_start_time = token.start_time	×
437
438	word = word + token.text	×
439	# Word boundary is either a space or the last character in the array
440	if token.text == " " or i == len(metadata.tokens) - 1:	×
441	word_duration = token.start_time - word_start_time	×
442
443	if word_duration < 0:	×
444	word_duration = 0	×
445
446	each_word = dict()	×
447	each_word["word"] = word	×
448	each_word["start_time"] = round(word_start_time, 4)	×
449	each_word["duration"] = round(word_duration, 4)	×
450
451	word_list.append(each_word)	×
452	# Reset
453	word = ""	×
454	word_start_time = 0	×
455
456	return word_list	×

EsupPortail / Esup-Pod / 5462629459

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous