Pre-built APKs
Links for pre-built APKs can be found in the following table:
Hint
It runs locally, without internet connection.
Chinese users |
URL |
|
Streaming speech recognition |
||
Simulated streaming
speech recognition
|
https://k2-fsa.github.io/sherpa/onnx/android/apk-simulate-streaming-asr.html |
|
Simulated streaming
speech recognition
with
Qualcomm NPUusing
QNN |
https://k2-fsa.github.io/sherpa/onnx/android/apk-qnn-simulate-streaming-asr.html |
|
Text-to-speech engine |
||
Text-to-speech |
||
Voice activity detection
(VAD)
|
||
VAD + non-streaming
speech recognition
|
||
Two-pass
speech recognition
|
||
Audio tagging |
||
Audio tagging (WearOS) |
https://k2-fsa.github.io/sherpa/onnx/audio-tagging/apk-wearos.html |
|
Speaker identification |
https://k2-fsa.github.io/sherpa/onnx/speaker-identification/apk.html |
|
Spoken language
identification
|
https://k2-fsa.github.io/sherpa/onnx/spoken-language-identification/apk.html |
|
Keyword spotting |
Note
Simulated streaming speech recognition: It uses a non-streaming ASR model for real-time/streaming speech recognition. For instance, use Whisper, Moonshine, SenseVoice, nvidia/parakeet-tdt-0.6b-v2, etc, for real-time speech recognition.