Google Cloud Speech Recognition Unity

# Google-Cloud-Speech-Recognition-Unity
This is a cross-platform speech recognition solution for Unity, powered by the Google Cloud Speech-to-Text API. It supports multiple platforms such as Android, iOS, Windows, and macOS, using a custom-built DLL to ensure multiplatform compatibility.

---

 

## Key Features

- Full integration with Google Cloud Speech-to-Text API
- Works on Android, iOS, Windows, macOS
- Custom DLL for multiplatform support
- Optional runtime voice detection
- Support for multiple microphones with device selection
- Context-aware recognition using Speech Contexts
- Handles long-running recognition operations
- Save recorded audio clips as `.wav` files
- Compatible with IL2CPP and Mono
- Partial support for WebGL
- Supports Unity Cloud Build

 

---

## Getting Started

### Recording Setup

In the `StartRecord` handler, call the `StartRecord(false)` function.  
Passing `false` disables runtime voice detection.  
To enable automatic background noise detection, use the `DetectThreshold` function.

### Speech Recognition Flow

1. Create a `RecognitionConfig` instance using `RecognitionConfig.GetDefault()`.
2. Modify parameters such as:
   - Language (from language dropdown)
   - Number of sound channels
   - Speech contexts for improving recognition accuracy
3. Create an array of `SpeechContext` objects to tailor recognition results to your context.
4. Prepare the audio for recognition:
   - Use `RecognitionAudioContent`
   - Convert the audio clip to base64 and assign to `content`
5. Create a `GeneralRecognitionRequest` with the config and audio data.
6. Call the `Recognize()` function with the request.

### Microphone Setup

The plugin provides these helper functions:

- `HasConnectedMicrophoneDevices()` – Checks if any microphones are available
- `SetMicrophoneDevice(string deviceName)` – Selects a specific microphone
- `GetMicrophoneDevices()` – Retrieves a list of all connected microphones

---

## Handling Recognition Results

You can log and handle events using simple handlers.  
The recognition success event provides a response object:

- Loop through `response.results`
- Each result contains `alternatives`
- Each alternative contains:
  - `transcript`: the recognized text
  - `words`: each word with timing metadata

   

Aggregate these to form your final recognized output.

---

## Setting Up Google Cloud

1. Go to: [Google Cloud Console](https://console.cloud.google.com/)
2. Create a new project or select an existing one.
3. Navigate to **Menu > API & Services > Credentials**
4. Click **Create Credentials > API Key**
5. Copy this key and paste it into the `GCSpeechRecognition` prefab in your Unity scene.
6. Go to the **APIs** section and enable the **Speech-to-Text API**.

Now your project is fully configured.

---

## Special Functions

| Function                        | Description                                                       |
|-------------------------------|-------------------------------------------------------------------|
| `GetOperation`                 | Gets info about a long-running operation                          |
| `GetListOperations`           | Lists all long-running operations                                 |
| `CancelAllRequests`           | Cancels all active recognition requests                           |
| `ReadyToRecord`               | Checks if recording is possible at the current moment             |
| `SaveLastRecordedAudioClip`   | Saves the last recording as a `.wav` file                         |
| `LongRunningRecognize`        | Starts a long-running recognition request                         |
| `CancelRequest(id)`           | Cancels a specific request by its ID                              |
| `HasMicrophonePermission`     | Checks if microphone access is granted                            |
| `RequestMicrophonePermission` | Requests permission for microphone access on platforms like Android|

---

## Notes

- Fully compatible with IL2CPP and Mono
- Unity Cloud Build support included
- Partial support for WebGL platforms

---

## License

This tool uses Google Cloud APIs under their respective licenses.  
Please ensure you follow Google’s [Terms of Use](https://cloud.google.com/terms) when using their services.