Initial commit of snowboy add-on

snowboy/Dockerfile

@@ -0,0 +1,20 @@
+FROM python:3.6
+LABEL maintainer="Michael Hansen <[email protected]>"
+
+RUN apt-get update && apt-get install -y python3-pip python3-dev \
+    jq \
+    libasound2-dev swig \
+    portaudio19-dev \
+    libatlas-base-dev \
+    libnanomsg-dev
+
+COPY snowboy-1.3.0.tar.gz /
+COPY nanomsg-python-master.zip /
+
+RUN pip3 install --no-cache-dir wheel
+RUN pip3 install --no-cache-dir /snowboy-1.3.0.tar.gz
+RUN pip3 install --no-cache-dir /nanomsg-python-master.zip
+
+COPY main.py /
+COPY run.sh /
+ENTRYPOINT ["/run.sh"]

snowboy/Makefile

@@ -0,0 +1,2 @@
+all:
+	docker build . -t synesthesiam/snowboy:1.3.0

snowboy/README.md

@@ -0,0 +1,30 @@
+Snowboy Wake Listener
+=========================
+
+Small service that listens for a wake word with [snowboy](https://snowboy.kitt.ai).
+Audio data is streamed in from [Rhasspy](https://github.com/synesthesiam/rhasspy-hassio-addon) via [nanomsg](https://nanomsg.org).
+
+
+Building
+----------
+
+To build the Docker image, run `make docker` in the project root.
+
+To create a local virtual environment, run the `create-venv.sh` shell script (expects a Debian distribution).
+
+
+Running
+---------
+
+To run with Docker:
+
+    docker run -it --network host synesthesiam/snowboy:1.3.0
+
+To run in a virtual environement (after running `create-venv.sh`):
+
+    ./run-venv.sh
+
+This will connect to ports 5000 (PUB) and 5001 (PULL) on localhost. By default, the hotword is "snowboy".
+
+Passing `--feedback` will let you see when audio data is being received and when the hotword is detected.
+See `--help` for additional command-line arguments.

snowboy/config.json

@@ -0,0 +1,22 @@
+{
+    "name": "Snowboy Wake System for Rhasspy",
+    "slug": "snowboy-rhasspy",
+    "version": "1.3.0-1",
+    "description": "Snowboy wake word detection (Kitt.ai) for Rhasspy voice assistant",
+    "startup": "application",
+    "boot": "auto",
+    "options": {
+        "pub_address": "tcp://75f2ff60-rhasspy:5000",
+        "pull_address": "tcp://75f2ff60-rhasspy:5001",
+        "model": "snowboy",
+        "sensitivity": 0.5,
+        "audio_gain": 1.0
+    },
+    "schema": {
+        "model": "str",
+        "sensitivity": "float",
+        "audio_gain": "float",
+        "pub_address": "str",
+        "pull_address": "str"
+    }
+}

snowboy/main.py

@@ -0,0 +1,123 @@
+#!/usr/bin/env python3
+import os
+import argparse
+import subprocess
+import logging
+logging.basicConfig(level=logging.DEBUG)
+
+from snowboy import snowboydetect, snowboydecoder
+from nanomsg import Socket, SUB, SUB_SUBSCRIBE, PUSH
+
+def main():
+    # Find available universal models (.umdl)
+    resource_dir = os.path.dirname(snowboydecoder.RESOURCE_FILE)
+    umdl_dir = os.path.join(resource_dir, 'models')
+
+    umdl_models = {
+        os.path.splitext(name)[0]: os.path.join(umdl_dir, name)
+        for name in os.listdir(umdl_dir)
+    }
+
+    # Parse arguments
+    parser = argparse.ArgumentParser(description='snowboy')
+    parser.add_argument('--pub-address',
+                        help='nanomsg address of PUB socket (default=tcp://127.0.0.1:5000)',
+                        type=str, default='tcp://127.0.0.1:5000')
+
+    parser.add_argument('--pull-address',
+                        help='nanomsg address of PULL socket (default=tcp://127.0.0.1:5001)',
+                        type=str, default='tcp://127.0.0.1:5001')
+
+    parser.add_argument('--payload', help='Payload string to send when wake word is detected (default=OK)',
+                        type=str, default='OK')
+
+    parser.add_argument('--model',
+                        action='append',
+                        type=str,
+                        help='Path to snowboy model file or one of %s (default=snowboy)' % list(umdl_models.keys()),
+                        default=[])
+
+    parser.add_argument('--sensitivity',
+                        action='append',
+                        help='Model sensitivity (default=0.5)',
+                        type=float,
+                        default=[])
+
+    parser.add_argument('--gain', help='Audio gain (default=1.0)',
+                        type=float, default=1.0)
+
+    parser.add_argument('--feedback', help='Show printed feedback', action='store_true')
+    args = parser.parse_args()
+
+    if len(args.model) == 0:
+        args.model = ['snowboy']
+
+    if len(args.sensitivity) == 0:
+        args.sensitivity = [0.5]
+
+    logging.debug(args)
+
+    # Create detector(s)
+    detectors = []
+
+    for i, model in enumerate(args.model):
+        model_path = umdl_models.get(model, model)
+        detector = snowboydetect.SnowboyDetect(
+            snowboydecoder.RESOURCE_FILE.encode(), model_path.encode())
+
+        if len(args.sensitivity) > i:
+            sensitivity_str = str(args.sensitivity[i]).encode()
+            detector.SetSensitivity(sensitivity_str)
+
+        detector.SetAudioGain(args.gain)
+        detectors.append(detector)
+
+    # Do detection
+    try:
+        payload = args.payload.encode()
+        first_frame = False
+
+        # Receive raw audio data via nanomsg
+        with Socket(SUB) as sub_socket:
+            sub_socket.connect(args.pub_address)
+            sub_socket.set_string_option(SUB, SUB_SUBSCRIBE, '')
+            logging.info('Connected to PUB socket at %s' % args.pub_address)
+
+            with Socket(PUSH) as push_socket:
+                # Response is sent via nanomsg
+                push_socket.connect(args.pull_address)
+                logging.info('Connected to PULL socket at %s' % args.pull_address)
+
+                while True:
+                    data = sub_socket.recv()  # audio data
+                    if args.feedback:
+                        print('.', end='', flush=True)
+
+                    if not first_frame:
+                        logging.debug('Receiving audio data from Rhasspy')
+                        first_frame = True
+
+                    for detector in detectors:
+                        index = detector.RunDetection(data)
+                        # Return is:
+                        # -2 silence
+                        # -1 error
+                        #  0 voice
+                        #  n index n-1
+                        if index > 0:
+                            # Hotword detected
+                            if args.feedback:
+                                print('!', end='', flush=True)
+
+                            logging.info('Wake word detected!')
+
+                            push_socket.send(payload)  # response
+                            first_frame = False
+
+    except KeyboardInterrupt:
+        pass
+
+# -----------------------------------------------------------------------------
+
+if __name__ == '__main__':
+    main()

snowboy/run.sh

@@ -0,0 +1,17 @@
+#!/usr/bin/env bash
+DIR="$( cd "$( dirname "$0" )" && pwd )"
+CONFIG_PATH="/data/options.json"
+
+pub_address="$(jq --raw-output '.pub_address' $CONFIG_PATH)"
+pull_address="$(jq --raw-output '.pull_address' $CONFIG_PATH)"
+model="$(jq --raw-output '.model' $CONFIG_PATH)"
+sensitivity="$(jq --raw-output '.sensitivity' $CONFIG_PATH)"
+audio_gain="$(jq --raw-output '.audio_gain' $CONFIG_PATH)"
+
+cd "$DIR"
+python3 main.py \
+        --pub-address "$pub_address" \
+        --pull-address "$pull_address" \
+        --model "$model" \
+        --sensitivity "$sensitivity" \
+        --gain "$audio_gain"