dgs.utils.track.Track

class dgs.utils.track.Track(N: int, curr_frame: int, states: list[State] | None = None, tid: int = -1)

A Track is a single (de-)queue containing multiple State’s, keeping the last N states.

Parameters:

• N – The max length of this track.

• states – A list of State objects, describing the initial values of this track. Default None.

• tid – The Track ID of this object. Default -1.

Methods

__init__(N: int, curr_frame: int, states: list[State] | None = None, tid: int = -1) → None

age(curr_frame: int) → int

Get the age of this track (in frames).

The age does not account for frames where the track has been deleted.

append(state: State) → None

Append a new state to the Track.

clear() → None

Clear all the states from the Track.

copy() → Track

Return a (deep) copy of self.

get_all() → State

Get all the states from the Track and stack them into a single State.

set_active() → None

set_inactive() → None

Set the status of this Track to inactive and clean up older states.

set_reactivated(tid: int) → None

set_removed() → None

set_status(status: TrackStatus, tid: int = 0) → None

Set the status of this Track.

to(*args, **kwargs) → Track

Call .to() like you do with any other torch.Tensor.

Attributes

N	Get the max length of this Track.
device	Get the device of every tensor in this Track.
id
nof_active
status