By name: whisper-realtime-stt description: Deploy OpenAI Whisper on NVIDIA Jetson Orin for real-time speech-to-text. Clones the deployment repo, installs dependencies including ffmpeg, tests the environment, and runs real-time STT from a USB microphone (e.g. reSpeaker). Includes Riva vs Whisper comparison context.
Real-Time Speech-to-Text with Whisper on Jetson Orin
Deploy Whisper on Jetson Orin for real-time speech-to-text processing directly on-device, eliminating network dependency and enhancing privacy. Uses a USB microphone for audio input.
Execution model
Run one phase at a time. After each phase:
- Relay all output to the user.
- If output contains
[STOP]→ stop immediately, consult the failure decision tree. - If output ends with
[OK]→ tell the user "Phase N complete" and proceed to the next phase.
Prerequisites
| Requirement | Detail |
|---|---|
| Jetson device | reComputer or other Jetson Orin-based device |
| Microphone | reSpeaker Mic Array v2.0 or other USB microphone |
| JetPack | With CUDA support |
| Network | Internet access for cloning repo and installing packages |
Phase 1 — Install dependencies (~5 min)
git clone https://github.com/LJ-Hao/Deploy-Whisper-on-NVIDIA-Jetson-Orin-for-Real-time-Speech-to-Text.git
cd Deploy-Whisper-on-NVIDIA-Jetson-Orin-for-Real-time-Speech-to-Text
pip install -r requirements.txt
sudo apt update && sudo apt install ffmpeg
Configure the microphone sample rate:
arecord -D hw:2,0 --dump-hw-params
[OK] when all packages install and ffmpeg is available. [STOP] if pip or apt install fails.
Phase 2 — Test environment (~1 min)
python test.py
Verify ffmpeg is installed:
ffmpeg -version
[OK] when test.py prints successful library import messages and ffmpeg -version shows version info. [STOP] if imports fail or ffmpeg is not found.
Phase 3 — Run real-time speech-to-text
python main.py
Speak into the microphone and observe real-time transcription output.
[OK] when transcription appears in the terminal as you speak. [STOP] if audio device errors or model loading fails.
Failure decision tree
| Symptom | Action |
|---|---|
pip install -r requirements.txt fails |
Check Python version ≥ 3.8. Try pip install --upgrade pip first. |
ffmpeg not found after install |
Run sudo apt install ffmpeg again. Verify with which ffmpeg. |
arecord — no soundcard found |
Check USB microphone connection. Run arecord -l to list devices. Adjust device ID (hw:X,0). |
test.py import errors |
Re-run pip install -r requirements.txt. Check for missing system libraries. |
main.py — CUDA out of memory |
Close other GPU processes. Use a smaller Whisper model variant. |
main.py — no audio input |
Verify microphone with arecord -D hw:2,0 -f S16_LE -r 16000 -d 5 test.wav. |
| Poor transcription accuracy | Ensure microphone sample rate is 16000 Hz. Reduce background noise. |
Reference files
references/source.body.md— Full Seeed Wiki tutorial with hardware setup photos, environment test screenshots, Riva vs Whisper comparison video, and project outlook (reference only)