Skip to content

Update transcriber.py #1344

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Update transcriber.py #1344

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
52 changes: 39 additions & 13 deletions python/vosk/transcriber/transcriber.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,7 @@
import datetime
import shlex
import subprocess

from tqdm.auto import tqdm
from vosk import KaldiRecognizer, Model
from queue import Queue
from timeit import default_timer as timer
Expand All @@ -22,15 +22,36 @@ def __init__(self, args):
self.args = args
self.queue = Queue()

def recognize_stream(self, rec, stream):
def recognize_stream(self, rec, stream, duration):

tot_samples = 0
result = []

show_bar=self.args.show_progress_bar

progress_bar = tqdm(
total=int(duration * SAMPLE_RATE),
unit="samples",
unit_scale=True,
unit_divisor=SAMPLE_RATE,
desc="Processing",
disable=not show_bar
)



while True:
data = stream.stdout.read(CHUNK_SIZE)


if len(data) == 0:
progress_bar.update(int(duration))
break
progress_bar.update((len(data) // 2) )





tot_samples += len(data)
if rec.AcceptWaveform(data):
Expand All @@ -42,9 +63,11 @@ def recognize_stream(self, rec, stream):
if jres["partial"] != "":
logging.info(jres)

progress_bar.close()
jres = json.loads(rec.FinalResult())
result.append(jres)


return result, tot_samples

async def recognize_stream_server(self, proc):
Expand Down Expand Up @@ -99,7 +122,7 @@ def format_result(self, result, words_per_line=7):
monologues = {"schemaVersion":"2.0", "monologues":[], "text":[]}
for part in result:
if part["text"] != "":
monologues["text"] += part["text"]
monologue["text"] += part["text"]
for _, res in enumerate(result):
if not "result" in res:
continue
Expand Down Expand Up @@ -133,12 +156,6 @@ async def server_worker(self):
start_time = timer()
proc = await self.resample_ffmpeg_async(input_file)
result, tot_samples = await self.recognize_stream_server(proc)
await proc.wait()

# Bad input, continue
if tot_samples == 0:
self.queue.task_done()
continue

processed_result = self.format_result(result)
if output_file != "":
Expand All @@ -148,6 +165,8 @@ async def server_worker(self):
else:
print(processed_result)

await proc.wait()

elapsed = timer() - start_time
logging.info("Execution time: {:.3f} sec; "\
"xRT {:.3f}".format(elapsed, float(elapsed) * (2 * SAMPLE_RATE) / tot_samples))
Expand All @@ -157,6 +176,15 @@ def pool_worker(self, inputdata):
logging.info("Recognizing {}".format(inputdata[0]))
start_time = timer()

# Get the duration of the input file in seconds
duration = float(
subprocess.check_output(
["ffprobe", "-v", "quiet", "-print_format", "compact=print_section=0:nokey=1:escape=csv",
"-show_entries", "format=duration", inputdata[0]]
).decode("utf-8").strip()
)
total_samples = int(duration * SAMPLE_RATE)

try:
stream = self.resample_ffmpeg(inputdata[0])
except FileNotFoundError as e:
Expand All @@ -168,11 +196,9 @@ def pool_worker(self, inputdata):

rec = KaldiRecognizer(self.model, SAMPLE_RATE)
rec.SetWords(True)
result, tot_samples = self.recognize_stream(rec, stream)
if tot_samples == 0:
return

result, tot_samples = self.recognize_stream(rec, stream,duration)
processed_result = self.format_result(result)

if inputdata[1] != "":
logging.info("File {} processing complete".format(inputdata[1]))
with open(inputdata[1], "w", encoding="utf-8") as fh:
Expand Down