Sonus
A Node.js library for adding voice interfaces with offline hotword detection and cloud speech recognition.
What is Sonus?
Sonus is a Node.js library for adding voice user interfaces (VUI) to hardware or software projects. It provides offline hotword detection using Snowboy, then streams audio to cloud speech recognition services like Google Cloud Speech for real-time transcription. It simplifies building always-listening voice interfaces similar to Alexa or Google Assistant.
Node.js developers building voice-controlled applications, IoT devices, or home automation systems that require offline wake-word detection and cloud-based speech recognition. It's particularly suited for projects on Linux (including Raspberry Pi) and macOS.
Developers choose Sonus for its simplicity and modular approach, bridging offline hotword detection with cloud speech APIs without complex setup. Its lightweight design and support for custom hotwords via Snowboy make it ideal for prototyping and deploying voice interfaces quickly.
Overview
:speech_balloon: /so.nus/ STT (speech to text) for Node with offline hotword detection
Use Cases
Best For
- Adding always-listening voice commands to Raspberry Pi projects without constant internet connectivity.
- Building custom voice assistants with personalized wake words for home automation systems.
- Prototyping voice interfaces for IoT devices that require real-time speech-to-text from cloud services.
- Creating voice-controlled software applications on Linux or macOS with offline hotword detection.
- Streaming audio to Google Cloud Speech API after detecting a custom hotword in Node.js environments.
- Developing educational or hobbyist projects that mimic commercial voice assistants like Alexa or Siri.
Not Ideal For
- Projects targeting Windows environments, as Sonus explicitly lacks Windows support.
- Applications requiring speech recognition from multiple cloud services beyond Google Cloud Speech, since only that is implemented.
- Teams needing fully offline speech-to-text capabilities, as Sonus relies on cloud APIs for transcription after hotword detection.
- Developers seeking a library with minimal external dependencies, due to required SoX installation and Snowboy integration.
Pros & Cons
Pros
Integrates Snowboy for customizable hotword detection without internet, allowing personalized wake words via online training, as shown in the README's hotword setup.
Simplifies streaming to Google Cloud Speech API with minimal code, enabling real-time transcription results after hotword triggers, evidenced by the concise usage examples.
Specifically supports Linux including Raspbian, making it ideal for IoT and hardware projects on affordable platforms like Raspberry Pi, as highlighted in the platform support.
Offers a dead-simple API that bridges hotword detection and speech recognition without bloat, allowing quick prototyping, as demonstrated in the 'create your own Alexa in less than a tweet' snippet.
Cons
Only Google Cloud Speech is fully supported as a streaming service; other listed options like Alexa or Wit.ai are marked as not implemented, restricting flexibility.
Requires SoX installation for audio handling and relies on Snowboy for hotwords, which needs online training and adds setup complexity, as noted in the dependencies section.
Explicitly does not support Windows, limiting cross-platform development and excluding a large segment of the developer community.
The project's dependencies and lack of recent updates suggested by build badges may lead to compatibility issues with newer Node.js versions or cloud API changes.
Frequently Asked Questions
Related Projects
AdGuard HomeNetwork-wide ads & trackers blocking DNS server
Sonic PiCode. Music. Live.
Pi-KVMOpen and inexpensive DIY IP-KVM based on Raspberry Pi
PoisonTapExploits locked/password protected computers over USB, drops persistent WebSocket-based backdoor, exposes internal router, and siphons cookies using Raspberry Pi Zero & Node.js.
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.