Document wake word detection entity (#1887)

docs/core/entity/stt.md

@@ -26,7 +26,7 @@ Properties should always only return information from memory and not do I/O (lik
 
 ### Process audio stream
 
-The process audio stream method is used send audio to an STT service and get text back.
+The process audio stream method is used to send audio to an STT service and get text back.
 
 ```python
 class MySpeechToTextEntity(SpeechToTextEntity):

docs/core/entity/wake_word.md

@@ -0,0 +1,50 @@
+---
+title: Wake Word Detection Entity
+sidebar_label: Wake Word Detection
+---
+
+A wake word detection entity allows other integrations or applications to detect wake words (also called hotwords) in an audio stream.
+
+A wake word detection entity is derived from the  [`homeassistant.components.wake_word.WakeWordDetectionEntity`](https://github.com/home-assistant/core/blob/dev/homeassistant/components/wake_word/__init__.py).
+
+## Properties
+
+:::tip
+Properties should always only return information from memory and not do I/O (like network requests).
+:::
+
+| Name                 | Type           | Default      | Description                                                                                                                 |
+|----------------------|----------------|--------------|-----------------------------------------------------------------------------------------------------------------------------|
+| supported_wake_words | list[WakeWord] | **Required** | The supported wake words of the service with:<ul><li>ww_id - unique identifier</li><li>name - human-readable name</li></ul> |
+
+## Methods
+
+### Process audio stream
+
+The process audio stream method is used to detect wake words. It must return a `DetectionResult` or `None` if the audio stream ends without a detection.
+
+```python
+class MyWakeWordDetectionEntity(WakeWordDetectionEntity):
+    """Represent a Wake Word Detection entity."""
+
+    async def async_process_audio_stream(
+        self, stream: AsyncIterable[tuple[bytes, int]]
+    ) -> DetectionResult | None:
+        """Try to detect wake word(s) in an audio stream with timestamps.
+
+        Audio must be 16Khz sample rate with 16-bit mono PCM samples.
+        """
+```
+
+The audio stream is made of tuples with the form `(timestamp, audio_chunk)` where:
+
+- `timestamp` is the number of milliseconds since the start of the audio stream
+- `audio_chunk` is a chunk of 16-bit signed mono PCM samples at 16Khz
+
+If a wake word is detected, a `DetectionResult` is returned with:
+
+- `ww_id` - the unique identifier of the detected wake word
+- `timestamp` - the timestamp of the audio chunk when detection occurred
+- `queued_audio` - optional audio chunks that will be forwarded to speech-to-text (see below)
+
+In an [Assist pipeline](/docs/voice/pipelines), the audio stream is shared between wake word detection and speech-to-text. This means that any audio chunk removed during wake word detection **can not be processed** by speech-to-text unless passed back in the `queued_audio` of a `DetectionResult`.