Examples for using vosk with docker.

Minimal example that prints out usage of the VOSK API. No other functionaliy.

To build the example, change into the directory and check the Dockerfile comments for build instructions.

Combines the open source "dlabpro" speech recognition system with the VOSK API to create a recognition system with simple (explicit or statistical) grammar. Models still need to be provided externally.

To build the example, change into the directory and check the Dockerfile comments for build instructions.

This implementation uses common code, so it needs to be built from the root directory:

docker build -f vosk_server_recikts/Dockerfile --progress=plain -t vosk_server_recikts .

Check the Dockerfile comments for more build options.

This implementation uses common code, so it needs to be built from the root directory:

docker build -f vosk_server_whisper/Dockerfile --progress=plain -t vosk_server_whisper .

This implementation uses a finetuned wav2vec2 model plus an n-gram language model.

To build the example, change into the directory and build the container from there.

cd vosk_server_wav2vec2
docker build --progress=plain -t vosk_server_wav2vec2 .

• Dr. Frank Duckhorn (Fraunhofer Institute for Ceramic Technologies and Systems IKTS, Dresden, Germany)

• Daniel Sobe (Foundation for the Sorbian people)

• Korla Baier

See file "LICENSE".