C++ version of openWakeWord.
- Download a release of the onnxruntime and extract to
lib/<arch>
where<arch>
isuname -m
. - Run
make
After building, run:
arecord -r 16000 -c 1 -f S16_LE -t raw - | \
build/openwakeword --model models/alexa_v0.1.onnx
You can add multiple --model <path>
arguments. See --help
for more options.