Splash is a Python-based voice assistant that listens for a wake word, processes voice commands, and responds with AI-generated answers. Designed for seamless interaction, Splash uses advanced speech recognition and natural-sounding text-to-speech (TTS) to bring conversations to life.
- Open in your browser Google AI Studio
- Create a key and copy it.
- Paste it in this line.
pip install -r requirements.txt- Microsoft Edge TTS: For high-quality text-to-speech synthesis.
- Faster Whisper: A fast and efficient library for transcribing speech using OpenAI's Whisper model.
- Run Splash
Execute the script, and Splash will actively listen for the wake word. The default wake word issplash. If it doesn't seem to work, head to Troubleshooting
- Wake Word Activation: Listens for a customizable wake word to activate.
- Speech-to-Text: Converts your speech to text using Faster Whisper.
- AI-Generated Responses: Interacts with you using the Gemini AI-powered response system.
- Text-to-Speech: Responds audibly using Edge TTS.
You can customize Splash by modifying the following:
- Wake Word: Change the
WAKE_WORDvariable in the script. - Input Device: Adjust
DEVICE_INDEXto use a specific audio input. - TTS Voice: Modify the voice in the
speak_textfunction (e.g.,en-GB-SoniaNeural), you can find them here.
- Microphone Issues: Ensure the correct
DEVICE_INDEXis set for your input device. - No Output Audio: Verify your speakers or headphones are selected as the output device.
- Wake Word Not Detected: Speak clearly and ensure the wake word matches what is set in the code.
- Dependencies Missing: Make sure all required libraries are installed.
Splash is open-source and licensed under the MIT License.