acies.vehicle_classifier.buffer

Classes

StreamBuffer	A buffer to hold streaming data.
TemporalEnsembleBuff	A buffer to hold temporal ensemble messages.

class acies.vehicle_classifier.buffer.StreamBuffer

Bases: object

A buffer to hold streaming data.

Parameters:

• size (int)

• _data (dict[str, dict[int, ndarray]])

• _meta (dict[str, dict[int, dict]])

• _timestamps (Counter[int])

__init__(size, _data=<factory>, _meta=<factory>, _timestamps=<factory>)

Parameters:

• size (int)

• _data (dict[str, dict[int, ndarray]])

• _meta (dict[str, dict[int, dict]])

• _timestamps (Counter[int])

add(topic, timestamp, samples, meta)

Add a new sample to the buffer.

Parameters:

• topic (str) – The topic of the sample (e.g., ‘rs1/mic’).

• timestamp (int) – The timestamp of the sample.

• samples (np.ndarray) – The audio samples.

• meta (dict) – Metadata associated with the samples.

get(keys, n)

Get n-second of samples for all keys

Parameters:

• keys (list[str]) – The keys to retrieve samples for.

• n (int) – The number of seconds of samples to retrieve.

Raises:

ValueError – If not enough data is available.

Returns:

The retrieved samples and their metadata.

Return type:

tuple[dict[str, dict[int, np.ndarray]], dict[str, dict[int, dict]]]

class acies.vehicle_classifier.buffer.TemporalEnsembleBuff

Bases: object

A buffer to hold temporal ensemble messages.

Parameters:

• buff_size (int)

• _data (dict[int, AciesMsg])

__init__(buff_size, _data=<factory>)

Parameters:

• buff_size (int)

• _data (dict[int, AciesMsg])

add(msg)

Add a new message to the buffer.

Parameters:

msg (AciesMsg) – The message to add.

ensemble(timestamp_now, ensemble_win_size, ensemble_size)

Perform temporal ensembling on the buffered messages.

Parameters:

• timestamp_now (int) – The current timestamp.

• ensemble_win_size (int) – The window size for ensembling.

• ensemble_size (int) – The number of messages to ensemble.

Raises:

ValueError – If not enough data is available for ensembling.

Returns:

The ensembled message.

Return type:

dict