vad.py

import webrtcvad
import wave
import collections
import contextlib
import sys
import glob
from os import walk
from os.path import join, basename, getctime
from json import dump, dumps
import csv
import time
import re
from itertools import takewhile


class Frame(object):
    """Represents a "frame" of audio data."""

    def __init__(self, bytes, timestamp, duration):
        self.bytes = bytes
        self.timestamp = timestamp
        self.duration = duration


def frame_generator(frame_duration_ms, audio, sample_rate):
    """Generates audio frames from PCM audio data.
    Takes the desired frame duration in milliseconds, the PCM data, and
    the sample rate.
    Yields Frames of the requested duration.
    """
    n = int(sample_rate * (frame_duration_ms / 1000.0) * 2)
    offset = 0
    timestamp = 0.0
    duration = (float(n) / sample_rate) / 2.0
    while offset + n < len(audio):
        yield Frame(audio[offset:offset + n], timestamp, duration)
        timestamp += duration
        offset += n


def read_wave(path):
    """Reads a .wav file.
    Takes the path, and returns (PCM audio data, sample rate).
    """
    with contextlib.closing(wave.open(path, 'rb')) as wf:
        sample_rate = wf.getframerate()
        sample_width = wf.getsampwidth()
        pcm_data = wf.readframes(wf.getnframes())
        num_channels = wf.getnchannels()
        assert num_channels == 1
        assert sample_width == 2
        assert sample_rate in (8000, 16000, 32000, 48000)
        return pcm_data, sample_rate, wf.getnframes()/float(sample_rate)


def vad_collector(sample_rate, frame_duration_ms,
                  padding_duration_ms, vad, frames):
    """Filters out non-voiced audio frames.
    Given a webrtcvad.Vad and a source of audio frames, yields only
    the voiced audio.
    Uses a padded, sliding window algorithm over the audio frames.
    When more than 90% of the frames in the window are voiced (as
    reported by the VAD), the collector triggers and begins yielding
    audio frames. Then the collector waits until 90% of the frames in
    the window are unvoiced to detrigger.
    The window is padded at the front and back to provide a small
    amount of silence or the beginnings/endings of speech around the
    voiced frames.
    Arguments:
    sample_rate - The audio sample rate, in Hz.
    frame_duration_ms - The frame duration in milliseconds.
    padding_duration_ms - The amount to pad the window, in milliseconds.
    vad - An instance of webrtcvad.Vad.
    frames - a source of audio frames (sequence or generator).
    Returns: A generator that yields PCM audio data.
    """
    num_padding_frames = int(padding_duration_ms / frame_duration_ms)
    # We use a deque for our sliding window/ring buffer.
    ring_buffer = collections.deque(maxlen=num_padding_frames)
    # We have two states: TRIGGERED and NOTTRIGGERED. We start in the
    # NOTTRIGGERED state.
    triggered = False

    voiced_frames = []
    for frame in frames:
        is_speech = vad.is_speech(frame.bytes, sample_rate)

        if not triggered:
            ring_buffer.append((frame, is_speech))
            num_voiced = len([f for f, speech in ring_buffer if speech])
            # If we're NOTTRIGGERED and more than 90% of the frames in
            # the ring buffer are voiced frames, then enter the
            # TRIGGERED state.
            if num_voiced > 0.9 * ring_buffer.maxlen:
                triggered = True
                # We want to yield all the audio we see from now until
                # we are NOTTRIGGERED, but we have to start with the
                # audio that's already in the ring buffer.
                for f, s in ring_buffer:
                    voiced_frames.append(f)
                ring_buffer.clear()
        else:
            # We're in the TRIGGERED state, so collect the audio data
            # and add it to the ring buffer.
            voiced_frames.append(frame)
            ring_buffer.append((frame, is_speech))
            num_unvoiced = len([f for f, speech in ring_buffer if not speech])
            # If more than 90% of the frames in the ring buffer are
            # unvoiced, then enter NOTTRIGGERED and yield whatever
            # audio we've collected.
            if num_unvoiced > 0.9 * ring_buffer.maxlen:
                triggered = False
                yield [f for f in voiced_frames]
                ring_buffer.clear()
                voiced_frames = []

    # If we have any leftover voiced audio when we run out of input,
    # yield it.
    if voiced_frames:
        yield [f for f in voiced_frames]


def collapse(table):
    byFn = {}
    for row in table:
        splitfn = row["Filename"].split("_")
        if len(splitfn) < 2:
            print(row["Filename"])
            continue
        if len(byFn.keys()) < 1 or splitfn[0] not in byFn:
            byFn[splitfn[0]] = {
                k: v for (k, v) in row.items() if k not in (
                    "Leading",
                    "Trailing"
                )
            }
            byFn[splitfn[0]]["irt"] = 0
            byFn[splitfn[0]]["Filename"] = splitfn[0]
        if "r1" in splitfn[1]:
            byFn[splitfn[0]]["irt"] += row["Trailing"]
        else:
            byFn[splitfn[0]]["irt"] += row["Leading"]
    return byFn


def main(args):

    try:
        fdir = args[0]
    except IndexError:
        fdir = "~/dissertation/recordings/16k/"

    main_start = time.time()
    fns = []
    rows = []
    looptimes = []
    fails = []
    # vad takes integer paramater 0-3 to indicate aggressiveness of ignoring
    # non-voice (0 is least. 3 is most agg)
    vad = webrtcvad.Vad(1)

    for root, dirs, files in walk(fdir):
        # build a list of recordings
        if "prac" in dirs:
            # eliminat practice items
            dirs.remove("prac")
        for name in files:
            if("wav" in name and "P" not in name):
                # We only want .wav files
                # Items prefaced with "P" are also practice
                fns.append(join(root, name))
    for fn in fns:
        # iterate through sound files and find speech blocks
        start = time.time()
        audio, sample_rate, length = read_wave(fn)
        bfn = basename(fn)
        # convert audio into chunks of a given length (in ms)
        frames = list(frame_generator(30, audio, sample_rate))
        # find the segments that are speech
        segments = list(vad_collector(sample_rate, 20, 300, vad, frames))
        # if we only found 1, we don't want it
        if len(segments) < 1:
            fails.append(fn)
            continue
        # extract participant and item from filename
        p, item = bfn.split(".", 1)
        item = "".join(takewhile(str.isdigit, item))
        rows.append({
            "Filename": bfn,
            # assume the leading silence ends when the first voiced segment
            # begins; also convert to ms
            "Leading": int(segments[0][0].timestamp * 1000),
            # assume the trailing silence ends when the last voiced segment
            # ends; also, convert to ms
            "Trailing": int((length - segments[-1][-1].timestamp) * 1000),
            "Length": int(length * 1000),  # full file duration
            "isFiller": "F" in bfn,
            "Condition_Q": "Q" in bfn,
            "Condition_GP": "Y" in bfn,
            "Participant": p,
            "Item": item
        })
        looptimes.append(time.time() - start)
    total_runtime = time.time() - main_start
    dts = time.strftime("%Y-%m-%d_%H%M%S")

    try:
        outf = args[1]
    except IndexError:
        outf = f"output-{dts}.csv"
    keys = rows[0].keys()
    with open(outf, "w") as outcsv:
        writer = csv.DictWriter(outcsv, fieldnames=keys)
        writer.writeheader()
        for row in rows:
            writer.writerow(row)

    ckeys = [field for field in keys if field not in (
        "Leading",
        "Trailing"
    )]
    ckeys.append("irt")
    with open("irt-"+outf, "w") as outirt:
        writer = csv.DictWriter(outirt, fieldnames=ckeys)
        writer.writeheader()
        for row in collapse(rows).values():
            writer.writerow(row)

    print(f"Processed {len(fns)} files in {(total_runtime * 1000):>.2f} ms " +
          f"with {len(fails)} failures. Longest iteration was " +
          f"{max(looptimes):>.4f} ms, shortest was {min(looptimes):>.4f} " +
          f"and ms average was {sum(looptimes)/len(looptimes):>.4f}")


if __name__ == '__main__':
    main(sys.argv[1:])
import webrtcvad
import wave
import collections
import contextlib
import sys
import glob
from os import walk
from os.path import join, basename, getctime
from json import dump, dumps
import csv
import time
import re
from itertools import takewhile


class Frame(object):
    """Represents a "frame" of audio data."""

    def __init__(self, bytes, timestamp, duration):
        self.bytes = bytes
        self.timestamp = timestamp
        self.duration = duration


def frame_generator(frame_duration_ms, audio, sample_rate):
    """Generates audio frames from PCM audio data.
    Takes the desired frame duration in milliseconds, the PCM data, and
    the sample rate.
    Yields Frames of the requested duration.
    """
    n = int(sample_rate * (frame_duration_ms / 1000.0) * 2)
    offset = 0
    timestamp = 0.0
    duration = (float(n) / sample_rate) / 2.0
    while offset + n < len(audio):
        yield Frame(audio[offset:offset + n], timestamp, duration)
        timestamp += duration
        offset += n


def read_wave(path):
    """Reads a .wav file.
    Takes the path, and returns (PCM audio data, sample rate).
    """
    with contextlib.closing(wave.open(path, 'rb')) as wf:
        sample_rate = wf.getframerate()
        sample_width = wf.getsampwidth()
        pcm_data = wf.readframes(wf.getnframes())
        num_channels = wf.getnchannels()
        assert num_channels == 1
        assert sample_width == 2
        assert sample_rate in (8000, 16000, 32000, 48000)
        return pcm_data, sample_rate, wf.getnframes()/float(sample_rate)


def vad_collector(sample_rate, frame_duration_ms,
                  padding_duration_ms, vad, frames):
    """Filters out non-voiced audio frames.
    Given a webrtcvad.Vad and a source of audio frames, yields only
    the voiced audio.
    Uses a padded, sliding window algorithm over the audio frames.
    When more than 90% of the frames in the window are voiced (as
    reported by the VAD), the collector triggers and begins yielding
    audio frames. Then the collector waits until 90% of the frames in
    the window are unvoiced to detrigger.
    The window is padded at the front and back to provide a small
    amount of silence or the beginnings/endings of speech around the
    voiced frames.
    Arguments:
    sample_rate - The audio sample rate, in Hz.
    frame_duration_ms - The frame duration in milliseconds.
    padding_duration_ms - The amount to pad the window, in milliseconds.
    vad - An instance of webrtcvad.Vad.
    frames - a source of audio frames (sequence or generator).
    Returns: A generator that yields PCM audio data.
    """
    num_padding_frames = int(padding_duration_ms / frame_duration_ms)
    # We use a deque for our sliding window/ring buffer.
    ring_buffer = collections.deque(maxlen=num_padding_frames)
    # We have two states: TRIGGERED and NOTTRIGGERED. We start in the
    # NOTTRIGGERED state.
    triggered = False

    voiced_frames = []
    for frame in frames:
        is_speech = vad.is_speech(frame.bytes, sample_rate)

        if not triggered:
            ring_buffer.append((frame, is_speech))
            num_voiced = len([f for f, speech in ring_buffer if speech])
            # If we're NOTTRIGGERED and more than 90% of the frames in
            # the ring buffer are voiced frames, then enter the
            # TRIGGERED state.
            if num_voiced > 0.9 * ring_buffer.maxlen:
                triggered = True
                # We want to yield all the audio we see from now until
                # we are NOTTRIGGERED, but we have to start with the
                # audio that's already in the ring buffer.
                for f, s in ring_buffer:
                    voiced_frames.append(f)
                ring_buffer.clear()
        else:
            # We're in the TRIGGERED state, so collect the audio data
            # and add it to the ring buffer.
            voiced_frames.append(frame)
            ring_buffer.append((frame, is_speech))
            num_unvoiced = len([f for f, speech in ring_buffer if not speech])
            # If more than 90% of the frames in the ring buffer are
            # unvoiced, then enter NOTTRIGGERED and yield whatever
            # audio we've collected.
            if num_unvoiced > 0.9 * ring_buffer.maxlen:
                triggered = False
                yield [f for f in voiced_frames]
                ring_buffer.clear()
                voiced_frames = []

    # If we have any leftover voiced audio when we run out of input,
    # yield it.
    if voiced_frames:
        yield [f for f in voiced_frames]


def collapse(table):
    byFn = {}
    for row in table:
        splitfn = row["Filename"].split("_")
        if len(splitfn) < 2:
            print(row["Filename"])
            continue
        if len(byFn.keys()) < 1 or splitfn[0] not in byFn:
            byFn[splitfn[0]] = {
                k: v for (k, v) in row.items() if k not in (
                    "Leading",
                    "Trailing"
                )
            }
            byFn[splitfn[0]]["irt"] = 0
            byFn[splitfn[0]]["Filename"] = splitfn[0]
        if "r1" in splitfn[1]:
            byFn[splitfn[0]]["irt"] += row["Trailing"]
        else:
            byFn[splitfn[0]]["irt"] += row["Leading"]
    return byFn


def main(args):

    try:
        fdir = args[0]
    except IndexError:
        fdir = "~/dissertation/recordings/16k/"

    main_start = time.time()
    fns = []
    rows = []
    looptimes = []
    fails = []
    # vad takes integer paramater 0-3 to indicate aggressiveness of ignoring
    # non-voice (0 is least. 3 is most agg)
    vad = webrtcvad.Vad(1)

    for root, dirs, files in walk(fdir):
        # build a list of recordings
        if "prac" in dirs:
            # eliminat practice items
            dirs.remove("prac")
        for name in files:
            if("wav" in name and "P" not in name):
                # We only want .wav files
                # Items prefaced with "P" are also practice
                fns.append(join(root, name))
    for fn in fns:
        # iterate through sound files and find speech blocks
        start = time.time()
        audio, sample_rate, length = read_wave(fn)
        bfn = basename(fn)
        # convert audio into chunks of a given length (in ms)
        frames = list(frame_generator(30, audio, sample_rate))
        # find the segments that are speech
        segments = list(vad_collector(sample_rate, 20, 300, vad, frames))
        # if we only found 1, we don't want it
        if len(segments) < 1:
            fails.append(fn)
            continue
        # extract participant and item from filename
        p, item = bfn.split(".", 1)
        item = "".join(takewhile(str.isdigit, item))
        rows.append({
            "Filename": bfn,
            # assume the leading silence ends when the first voiced segment
            # begins; also convert to ms
            "Leading": int(segments[0][0].timestamp * 1000),
            # assume the trailing silence ends when the last voiced segment
            # ends; also, convert to ms
            "Trailing": int((length - segments[-1][-1].timestamp) * 1000),
            "Length": int(length * 1000),  # full file duration
            "isFiller": "F" in bfn,
            "Condition_Q": "Q" in bfn,
            "Condition_GP": "Y" in bfn,
            "Participant": p,
            "Item": item
        })
        looptimes.append(time.time() - start)
    total_runtime = time.time() - main_start
    dts = time.strftime("%Y-%m-%d_%H%M%S")

    try:
        outf = args[1]
    except IndexError:
        outf = f"output-{dts}.csv"
    keys = rows[0].keys()
    with open(outf, "w") as outcsv:
        writer = csv.DictWriter(outcsv, fieldnames=keys)
        writer.writeheader()
        for row in rows:
            writer.writerow(row)

    ckeys = [field for field in keys if field not in (
        "Leading",
        "Trailing"
    )]
    ckeys.append("irt")
    with open("irt-"+outf, "w") as outirt:
        writer = csv.DictWriter(outirt, fieldnames=ckeys)
        writer.writeheader()
        for row in collapse(rows).values():
            writer.writerow(row)

    print(f"Processed {len(fns)} files in {(total_runtime * 1000):>.2f} ms " +
          f"with {len(fails)} failures. Longest iteration was " +
          f"{max(looptimes):>.4f} ms, shortest was {min(looptimes):>.4f} " +
          f"and ms average was {sum(looptimes)/len(looptimes):>.4f}")


if __name__ == '__main__':
    main(sys.argv[1:])