drop snowboy

master
Bel LaPointe 2023-11-30 09:01:50 -07:00
parent 3b4295d026
commit 827436d96c
15 changed files with 0 additions and 372 deletions

View File

@ -1,29 +0,0 @@
# stt
## listen on linux
https://wiki.archlinux.org/title/PulseAudio/Examples
```
10. ALSA monitor source
To be able to record from a monitor source (a.k.a. "What-U-Hear", "Stereo Mix"), use pactl list to find out the name of the source in PulseAudio (e.g. alsa_output.pci-0000_00_1b.0.analog-stereo.monitor). Then add lines like the following to /etc/asound.conf or ~/.asoundrc:
pcm.pulse_monitor {
type pulse
device alsa_output.pci-0000_00_1b.0.analog-stereo.monitor
}
ctl.pulse_monitor {
type pulse
device alsa_output.pci-0000_00_1b.0.analog-stereo.monitor
}
Now you can select pulse_monitor as a recording source.
Alternatively, you can use pavucontrol to do this: make sure you have set up the display to "All input devices", then select "Monitor of [your sound card]" as the recording source.
```
```bash
$ pactl list | grep -A 50 RUNNING | grep -E 'RUNNING|Name:|Monitor Source:' | grep Monitor.Source | head -n 1 | awk '{print $NF}'
```

View File

@ -1,99 +0,0 @@
#! /bin/bash
echo https://github.com/seasalt-ai/snowboy
cd "$(dirname "$(realpath "$BASH_SOURCE")")"
set -e
set -o pipefail
if [ ! -d ./snowboy.git.d ]; then
git clone https://github.com/seasalt-ai/snowboy snowboy.git.d
fi
timeout 2 docker version &> /dev/null
if ! docker images | grep snowboy-pmdl.*latest &> /dev/null; then
pushd snowboy.git.d
docker build -t snowboy-pmdl:latest .
popd
fi
export HOTWORD="${HOTWORD:-${TRAIN:-default_hotword}}"
if [ -n "$TRAIN" ] || [ ! -d ./model ] || [ ! -f ./model/$HOTWORD.pmdl ]; then
mkdir -p model
pushd model
rm -f ./record{1,2,3}.wav || true
echo "record 3 instances of '$HOTWORD'" >&2
for i in 1 2 3; do
read -p "[$i/3] ready? you get 3 seconds."
(
timeout 3 rec \
-r 16000 \
-c 1 \
-b 16 \
-e signed-integer \
-t wav \
record$i.wav
) || true
ls record$i.wav
done
popd
docker run \
--rm \
-it \
-v "$(realpath ./model)":/snowboy-master/examples/Python/model \
snowboy-pmdl:latest
mv ./model/hotword.pmdl ./model/$HOTWORD.pmdl
if [ -n "$TRAIN" ]; then
exit 0
fi
fi
if false; then
if ! which swig; then
brew install swig
fi
pip3 install pyaudio
pushd snowboy.git.d/swig/Python3/
make
popd
cd snowboy.git.d/examples/Python3/
echo '
import snowboydecoder
import datetime
detected_callback = lambda *args: print(datetime.datetime.now(), "GOTCHA")
d = snowboydecoder.HotwordDetector("../../../model/'"$HOTWORD"'.pmdl", sensitivity=0.5, audio_gain=1)
d.start(detected_callback)
' > breel.py
echo GO
cleanup() {
echo OK IM DONE NOW
}
trap cleanup EXIT
python3 ./breel.py
else
resources="$(realpath snowboy.git.d/resources/common.res)"
hotword="$(realpath ./model/$HOTWORD.pmdl)"
GOPROXY= go build -o snowboy
if [ -z "$PUSH" ]; then
./snowboy \
-ms "$hotword/$HOTWORD" \
-r "$resources" \
-s 0.5 \
"$@"
else
echo '
FROM registry-app.eng.qops.net:5001/imported/alpine:3.16
WORKDIR /main/
COPY ./snowboy.git.d/resources/common.res ./
COPY ./model/hotword.pmdl ./
COPY ./snowboy ./
ENTRYPOINT ["sh", "-c", "true; echo copying /main/ to /mnt/; cp /main/* /mnt/"]
CMD []
' > Dockerfile
docker build -t registry-app.eng.qops.net:5001/breel/snowboy:latest .
docker push registry-app.eng.qops.net:5001/breel/snowboy:latest
fi
fi

View File

@ -1,13 +0,0 @@
module snowboy
go 1.19
require (
github.com/brentnd/go-snowboy v0.0.0-20190301212623-e19133c572af
github.com/gordonklaus/portaudio v0.0.0-20221027163845-7c3b689db3cc
)
require (
github.com/Kitt-AI/snowboy v1.3.0 // indirect
github.com/stretchr/testify v1.8.1 // indirect
)

View File

@ -1,22 +0,0 @@
github.com/Kitt-AI/snowboy v1.3.0 h1:PjBVN84M/9tAzDBQXILAKMoJMxt/fT0nhJ1rhKtVRUc=
github.com/Kitt-AI/snowboy v1.3.0/go.mod h1:sDzzMXFQ1wFkXkZaX/ant0xJsizGVq/9hyKb7ZB3cNI=
github.com/brentnd/go-snowboy v0.0.0-20190301212623-e19133c572af h1:ijY5OHNQs3CdzTN2XT+zByIsR1QVyXTvOUSkQcBm6pw=
github.com/brentnd/go-snowboy v0.0.0-20190301212623-e19133c572af/go.mod h1:XcT4k8Tn9hrM5SLVvu5hNQbAC6GojXM0MXz1Rt8CL68=
github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
github.com/gordonklaus/portaudio v0.0.0-20221027163845-7c3b689db3cc h1:yYLpN7bJxKYILKnk20oczGQOQd2h3/7z7/cxdD9Se/I=
github.com/gordonklaus/portaudio v0.0.0-20221027163845-7c3b689db3cc/go.mod h1:WY8R6YKlI2ZI3UyzFk7P6yGSuS+hFwNtEzrexRyD7Es=
github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
github.com/stretchr/objx v0.4.0/go.mod h1:YvHI0jy2hoMjB+UWwv71VJQ9isScKT/TqJzVSSt89Yw=
github.com/stretchr/objx v0.5.0/go.mod h1:Yh+to48EsGEfYuaHDzXPcE3xhTkx73EhmCGUpEOglKo=
github.com/stretchr/testify v1.7.1/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
github.com/stretchr/testify v1.8.0/go.mod h1:yNjHg4UonilssWZ8iaSj1OCr/vHnekPRkoO+kdMU+MU=
github.com/stretchr/testify v1.8.1 h1:w7B6lhMri9wdJUVmEZPGGhZzrYTPvgJArz7wNPgYKsk=
github.com/stretchr/testify v1.8.1/go.mod h1:w2LPCIKwWwSfY2zedu0+kehJoqGctiVI29o6fzry7u4=
gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
gopkg.in/yaml.v3 v3.0.1 h1:fxVm/GzAzEWqLHuvctI91KS9hhNmmWOoWu0XTYJS7CA=
gopkg.in/yaml.v3 v3.0.1/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=

View File

@ -1,122 +0,0 @@
// This example streams the microphone thru Snowboy to listen for the hotword,
// by using the PortAudio interface.
//
// HOW TO USE:
// go run examples/Go/listen/main.go [path to snowboy resource file] [path to snowboy hotword file]
//
package main
import (
"bytes"
"encoding/binary"
"flag"
"fmt"
"log"
"path"
"strings"
"time"
"github.com/brentnd/go-snowboy"
"github.com/gordonklaus/portaudio"
)
// Sound represents a sound stream implementing the io.Reader interface
// that provides the microphone data.
type Sound struct {
stream *portaudio.Stream
data []int16
}
// Init initializes the Sound's PortAudio stream.
func (s *Sound) Init() {
inputChannels := 1
outputChannels := 0
sampleRate := 16000
s.data = make([]int16, 1024)
// initialize the audio recording interface
err := portaudio.Initialize()
if err != nil {
fmt.Errorf("Error initialize audio interface: %s", err)
return
}
// open the sound input stream for the microphone
stream, err := portaudio.OpenDefaultStream(inputChannels, outputChannels, float64(sampleRate), len(s.data), s.data)
if err != nil {
fmt.Errorf("Error open default audio stream: %s", err)
return
}
err = stream.Start()
if err != nil {
fmt.Errorf("Error on stream start: %s", err)
return
}
s.stream = stream
}
// Close closes down the Sound's PortAudio connection.
func (s *Sound) Close() {
s.stream.Close()
portaudio.Terminate()
}
// Read is the Sound's implementation of the io.Reader interface.
func (s *Sound) Read(p []byte) (int, error) {
s.stream.Read()
buf := &bytes.Buffer{}
for _, v := range s.data {
binary.Write(buf, binary.LittleEndian, v)
}
copy(p, buf.Bytes())
return len(p), nil
}
func main() {
resources := flag.String("r", "", "path to the .res file")
models := flag.String("ms", "", "comma delimited path to the .?mdl file/output")
sensitivity := flag.Float64("s", 0.45, "0..1")
quiet := flag.Bool("q", false, "emit '1' on detect else silent")
flag.Parse()
if *resources == "" || *models == "" {
panic("all flags must be set")
}
// open the mic
mic := &Sound{}
mic.Init()
defer mic.Close()
// open the snowboy detector
d := snowboy.NewDetector(*resources)
defer d.Close()
// set the handlers
for _, modelStrC := range strings.Split(*models, ",") {
modelStr := modelStrC
d.HandleFunc(snowboy.NewHotword(path.Dir(modelStr), float32(*sensitivity)), func(string) {
if !*quiet {
log.Println(path.Base(modelStr))
}
fmt.Println(path.Base(modelStr))
})
}
d.HandleSilenceFunc(1*time.Second, func(string) {
if !*quiet {
log.Println("...")
}
})
// display the detector's expected audio format
sr, nc, bd := d.AudioFormat()
log.Printf("sample rate=%d, num channels=%d, bit depth=%d\n", sr, nc, bd)
// start detecting using the microphone
d.ReadAndDetect(mic)
}

Binary file not shown.

Binary file not shown.

Binary file not shown.

Binary file not shown.

Binary file not shown.

Binary file not shown.

View File

@ -1,82 +0,0 @@
#! /bin/bash
main() {
cleanup() {
killall -9 $(jobs -p)
killall snowboy
}
trap cleanup EXIT
if [ ! -e /tmp/stt.fifo ]; then
mkfifo /tmp/stt.fifo
fi
echo starting in
for ((i=2; i>0; i--)); do
echo "...$i..."
sleep 1
done
local models=($(
cat pyautogui.yaml \
| gojq -r -c --yaml-input '
to_entries[] | "model/"+.key+".pmdl/"+.key
' \
| tr '\n' ',' \
| sed 's/,$//'
))
echo models=$models
./snowboy -r resources.res -ms $models "$@" > /tmp/stt.fifo &
python3 -c '
import pyautogui
import time
keys = set()
def toggle(key):
global keys
if key in keys:
release(key)
else:
hold(key)
def hold(key):
global keys
for keyin in [todrop for todrop in keys]:
if keyin != key:
release(keyin)
keys = set()
keys.add(key)
print()
print("pressing", key)
print()
pyautogui.keyDown(key)
def release(key):
print()
print("releasing", key)
print()
pyautogui.keyUp(key)
def main():
with open("/tmp/stt.fifo", "r") as q:
for line in q:
handle(line.strip())
import yaml
mapping = yaml.safe_load(open("./pyautogui.yaml", "r"))
print(mapping)
def handle(cmd):
global mapping
hold(mapping.get(cmd))
main()
'
}
if [ "$0" == "$BASH_SOURCE" ]; then
main "$@"
fi

View File

@ -1,5 +0,0 @@
up: w
down: s
left: a
right: d
jump: w

Binary file not shown.