mycroft-precise support

README.md

@@ -94,6 +94,7 @@ See `servers` section of `configuration.yaml` file.
 * wake 
     * [porcupine1](https://github.com/Picovoice/porcupine)
     * [precise-lite](https://github.com/mycroftAI/mycroft-precise)
+    * [mycroft-precise](https://github.com/mycroftAI/mycroft-precise)
     * [snowboy](https://github.com/Kitt-AI/snowboy)
 * vad
     * [silero](https://github.com/snakers4/silero-vad)

programs/wake/mycroft_precise/bin/precise_server.py

@@ -0,0 +1,146 @@
+#!/usr/bin/env python3
+import argparse
+import logging
+import os
+import socket
+from pathlib import Path
+
+from precise_runner import PreciseEngine, PreciseRunner, ReadWriteStream
+
+from rhasspy3.audio import AudioChunk, AudioStart, AudioStop
+from rhasspy3.event import read_event, write_event
+from rhasspy3.wake import Detection, NotDetected
+
+_FILE = Path(__file__)
+_DIR = _FILE.parent
+_LOGGER = logging.getLogger(_FILE.stem)
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser()
+    parser.add_argument("model", help="Location to .pb model file to use (with .pb.params)")
+    parser.add_argument(
+        "--socketfile", required=True, help="Path to Unix domain socket file"
+    )
+    parser.add_argument("--engine", default='.venv/bin/precise-engine', help="Path to precise-engine executable")
+    parser.add_argument("--chunk_size", type=int, default=2048, help="Number of *bytes* per prediction. Higher numbers decrease CPU usage but increase latency")
+    parser.add_argument("--trigger_level", type=int, default=3, help="Number of chunk activations needed to trigger detection event. Higher values add latency but reduce false positives")
+    parser.add_argument("--sensitivity", type=float, default=0.5, help="From 0.0 to 1.0, how sensitive the network should be")
+    parser.add_argument("--debug", action="store_true", help="Log DEBUG messages")
+    args = parser.parse_args()
+
+    logging.basicConfig(level=logging.DEBUG if args.debug else logging.INFO)
+
+    # Need to unlink socket if it exists
+    try:
+        os.unlink(args.socketfile)
+    except OSError:
+        pass
+
+    try:
+        # Create socket server
+        sock = socket.socket(socket.AF_UNIX, socket.SOCK_STREAM)
+        sock.bind(args.socketfile)
+        sock.listen()
+
+        # Load converted faster-whisper model
+        engine = PreciseEngine(args.engine, args.model, args.chunk_size)
+        stream = ReadWriteStream()
+
+        conn_file_s = None
+        is_detected = False
+        timestamp = None
+        name = os.path.basename(args.model).removesuffix(".pb")
+        sensitivity = args.sensitivity
+
+        def activation():
+            nonlocal is_detected, timestamp, conn_file_s, name
+
+            _LOGGER.info("Detected")
+
+            if is_detected:
+                return
+            if conn_file_s is None:
+                return
+
+            try:
+                write_event(
+                    Detection(
+                        name=name, timestamp=timestamp
+                    ).event(),
+                    conn_file_s,
+                )  # type: ignore
+
+                is_detected = True
+            except Exception:
+                pass
+
+        def prediction(prob):
+            nonlocal is_detected, conn_file_s
+
+            if not is_detected:
+                return
+            if conn_file_s is None:
+                return
+            if prob >= sensitivity:
+                return
+
+            try:
+                write_event(NotDetected().event(), conn_file_s)  # type: ignore
+            except Exception:
+                pass
+
+            is_detected = False
+
+        runner = PreciseRunner(engine, trigger_level=args.trigger_level, sensitivity=args.sensitivity, stream=stream, on_activation=activation, on_prediction=prediction)
+        runner.start()
+        _LOGGER.info("Ready")
+
+        # Listen for connections
+        while True:
+            try:
+                connection, client_address = sock.accept()
+                _LOGGER.debug("Connection from %s", client_address)
+
+                with connection, connection.makefile(mode="rwb") as conn_file:
+                    conn_file_s = conn_file
+                    while True:
+                        event = read_event(conn_file)  # type: ignore
+                        if event is None:
+                            break
+
+                        if AudioStart.is_type(event.type):
+                            _LOGGER.debug("Receiving audio")
+                            continue
+
+                        if AudioStop.is_type(event.type):
+                            _LOGGER.debug("Audio stopped")
+                            break
+
+                        if not AudioChunk.is_type(event.type):
+                            continue
+
+                        chunk = AudioChunk.from_event(event)
+
+                        timestamp = chunk.timestamp
+                        stream.write(chunk.audio)
+
+                    if is_detected:
+                        write_event(NotDetected().event(), conn_file)  # type: ignore
+
+            except KeyboardInterrupt:
+                break
+            except Exception:
+                _LOGGER.exception("Error communicating with socket client")
+            finally:
+                conn_file_s = None
+                is_detected = False
+    finally:
+        os.unlink(args.socketfile)
+        runner.stop()
+
+
+# -----------------------------------------------------------------------------
+
+if __name__ == "__main__":
+    main()

programs/wake/mycroft_precise/bin/precise_stream.py

@@ -0,0 +1,83 @@
+#!/usr/bin/env python3
+import argparse
+import logging
+import os
+from pathlib import Path
+
+from precise_runner import PreciseEngine, PreciseRunner, ReadWriteStream
+
+from rhasspy3.audio import AudioChunk, AudioStop
+from rhasspy3.event import read_event, write_event
+from rhasspy3.wake import Detection, NotDetected
+
+_FILE = Path(__file__)
+_DIR = _FILE.parent
+_LOGGER = logging.getLogger(_FILE.stem)
+
+# -----------------------------------------------------------------------------
+
+
+def main() -> None:
+    """Main method."""
+    parser = argparse.ArgumentParser()
+    parser.add_argument("model", help="Location to .pb model file to use (with .pb.params)")
+    parser.add_argument("--engine", default='.venv/bin/precise-engine', help="Path to precise-engine executable")
+    parser.add_argument("--chunk_size", type=int, default=2048, help="Number of *bytes* per prediction. Higher numbers decrease CPU usage but increase latency")
+    parser.add_argument("--trigger_level", type=int, default=3, help="Number of chunk activations needed to trigger detection event. Higher values add latency but reduce false positives")
+    parser.add_argument("--sensitivity", type=float, default=0.5, help="From 0.0 to 1.0, how sensitive the network should be")
+    parser.add_argument("--debug", action="store_true", help="Log DEBUG messages")
+    args = parser.parse_args()
+
+    logging.basicConfig(level=logging.DEBUG if args.debug else logging.INFO)
+
+    engine = PreciseEngine(args.engine, args.model, args.chunk_size)
+    stream = ReadWriteStream()
+
+    is_detected = False
+    timestamp = None
+    name = os.path.basename(args.model).removesuffix(".pb")
+
+    def activation():
+        nonlocal is_detected, timestamp, name
+        try:
+            write_event(
+                Detection(
+                    name=name, timestamp=timestamp
+                ).event()
+            )
+            is_detected = True
+        except Exception:
+            pass
+
+
+    runner = PreciseRunner(engine, trigger_level=args.trigger_level, sensitivity=args.sensitivity, stream=stream, on_activation=activation)
+    runner.start()
+
+    try:
+        while True:
+            event = read_event()
+            if event is None:
+                break
+
+            if AudioStop.is_type(event.type):
+                break
+
+            if not AudioChunk.is_type(event.type):
+                continue
+
+            chunk = AudioChunk.from_event(event)
+
+            timestamp = chunk.timestamp
+            stream.write(chunk.audio)
+
+        if is_detected:
+            write_event(NotDetected().event())
+    except KeyboardInterrupt:
+        pass
+
+    runner.stop()
+
+# -----------------------------------------------------------------------------
+
+if __name__ == "__main__":
+    main()

programs/wake/mycroft_precise/requirements.txt

@@ -0,0 +1 @@
+git+https://github.com/MycroftAI/mycroft-precise.git@refs/pull/242/merge

programs/wake/mycroft_precise/script/download.py

@@ -0,0 +1,61 @@
+#!/usr/bin/env python3
+import argparse
+import logging
+import re
+import subprocess
+import tarfile
+from pathlib import Path
+from urllib.request import urlopen
+
+_DIR = Path(__file__).parent
+_LOGGER = logging.getLogger("download")
+
+
+def get_models():
+    _LOGGER.info("Downloading list of models...")
+    url = "https://github.com/MycroftAI/Precise-Community-Data.git/branches/master"
+    url2 = "https://github.com/MycroftAI/Precise-Community-Data/raw/master/"
+    filelist = subprocess.run(["svn", "ls", "-R", url], capture_output=True)
+    result = {}
+    for file in [x.decode() for x in filelist.stdout.split(b'\n')]:
+        if not file.endswith('.tar.gz') or not "/models/" in file:
+            continue
+        result.update({re.split('[0-9]', file)[0].split('/')[-1].removesuffix('-'): url2 + file})
+    return result
+
+def main() -> None:
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--destination", help="Path to destination directory (default: data)"
+    )
+    parser.add_argument(
+        "--name",
+        help="Model name",
+    )
+    args = parser.parse_args()
+    logging.basicConfig(level=logging.INFO)
+
+    if args.destination:
+        args.destination = Path(args.destination)
+    else:
+        # Assume we're in programs/wake/mycroft_precise/script
+        data_dir = _DIR.parent.parent.parent.parent / "data"
+        args.destination = data_dir / "wake" / "mycroft_precise"
+
+    args.destination.parent.mkdir(parents=True, exist_ok=True)
+
+    models = get_models()
+    if args.name is None or args.name not in models:
+        _LOGGER.info("Available models: %s", list(models.keys()))
+        return
+
+    _LOGGER.info("Downloading %s", args.name)
+    _LOGGER.info("URL: %s", models[args.name])
+    with urlopen(models[args.name]) as response:
+        with tarfile.open(mode="r|*", fileobj=response) as tar_gz:
+            _LOGGER.info("Extracting to %s", args.destination)
+            tar_gz.extractall(args.destination)
+
+
+if __name__ == "__main__":
+    main()

programs/wake/mycroft_precise/script/server

@@ -0,0 +1,20 @@
+#!/usr/bin/env bash
+set -eo pipefail
+
+# Directory of *this* script
+this_dir="$( cd "$( dirname "$0" )" && pwd )"
+
+# Base directory of repo
+base_dir="$(realpath "${this_dir}/..")"
+
+# Path to virtual environment
+: "${venv:=${base_dir}/.venv}"
+
+if [ -d "${venv}" ]; then
+    source "${venv}/bin/activate"
+fi
+
+socket_dir="${base_dir}/var/run"
+mkdir -p "${socket_dir}"
+
+python3 "${base_dir}/bin/precise_server.py" --socketfile "${socket_dir}/precise.socket" "$@"

programs/wake/mycroft_precise/script/setup

@@ -0,0 +1,47 @@
+#!/usr/bin/env bash
+set -eo pipefail
+
+# Directory of *this* script
+this_dir="$( cd "$( dirname "$0" )" && pwd )"
+
+# Base directory of repo
+base_dir="$(realpath "${this_dir}/..")"
+
+# Path to virtual environment
+: "${venv:=${base_dir}/.venv}"
+
+# Python binary to use
+: "${PYTHON=python3}"
+
+python_version="$(${PYTHON} --version)"
+
+if [ ! -d "${venv}" ]; then
+    # Create virtual environment
+    echo "Creating virtual environment at ${venv} (${python_version})"
+    rm -rf "${venv}"
+    "${PYTHON}" -m venv "${venv}"
+    source "${venv}/bin/activate"
+
+    pip3 install --upgrade pip
+    pip3 install --upgrade wheel setuptools
+else
+    source "${venv}/bin/activate"
+fi
+
+# Install Python dependencies
+echo 'Installing Python dependencies'
+pip3 install -r "${base_dir}/requirements.txt"
+
+# Download official models
+URL="https://github.com/MycroftAI/precise-data.git/branches/models"
+DEST="$this_dir/../../../../data/wake/mycroft_precise"
+echo "Downloading models"
+mkdir -p "$DEST"
+for MODEL in $(svn ls "$URL" | grep '\.pb'); do
+  echo "$MODEL"
+  svn cat "$URL/$MODEL" > "$DEST/$MODEL"
+done
+
+# -----------------------------------------------------------------------------
+
+echo "OK"