类class agentscope.realtime

VADBase

class VADBase(ABC)

源码Source agentscope/realtime/_vad.py:15–48

语音活动检测基类，处理低延迟的实时语音与多模态交互。位于 agentscope.realtime 包，继承自 ABC。核心成员包括 push、reset。

Implementations are stateful: push is called with consecutive chunks and returns a transition only on the frame where it happens.

属性摘要Attribute Summary

属性Attribute	类型Type	说明Description	值Value
`sample_rate`	int	`sample_rate` 字段。	—

方法Method	说明Description
`def push(pcm: bytes) ->` `SpeechTransition \| None`abstractabstractmethod	设置推送。 Feed one chunk of audio and report a transition, if any.
`def reset() -> None`abstractabstractmethod	清空或重置内部状态。 Drop internal state. Called whenever the audio stream has a gap — a reconnect, or the microphone being unmuted — since the silence counters carried across it would be stale.

int

sample_rate 字段。声明为 int。

The sample rate field.

源码Source agentscope/realtime/_vad.py:22

def push(pcm: bytes) -> SpeechTransition | None

abstractabstractmethod

设置推送。

This is an edge detector, not a level query: it returns a value only on the chunk where speech starts or stops.

参数Parameters

名称Name	类型Type	说明Description
`pcm`	bytes	pcm（类型 `bytes`）。PCM16 mono audio at :attr:`sample_rate`.

返回Returns SpeechTransition | None

返回SpeechTransition|None。`SpeechTransition | None`:

源码Source agentscope/realtime/_vad.py:26–42

def reset() -> None

abstractabstractmethod

清空或重置内部状态。

Drop internal state.

返回Returns None

无返回值。No value is returned.

源码Source agentscope/realtime/_vad.py:45–48