Detections

byotrack.api.detector.detections.sorted_alphanumeric(data: Iterable[str])

Sorts alphanumeriacally an iterable of strings

“1” < “2” < “10” < “foo1” < “foo2” < “foo3”

byotrack.api.detector.detections.relabel_consecutive(segmentation: Tensor) → Tensor

Relabel a segmentation mask so that labels are consecutives

For N instances, labels are 0 for background and then [1:N] for each instance.

Parameters:

segmentation (torch.Tensor) – Segmentation mask Shape: (H, W), dtype: int32

Returns:

The same segmentation mask where labels are consecutive (from 0 to N)

Return type:

torch.Tensor

class byotrack.api.detector.detections.Detections(data: Dict[str, Tensor], frame_id: int = -1)

Bases: object

Detections for a given frame

Built from a data dict. The data has to contained one of “position”, “bbox” or “segmentation” keys that respectively define the positions of instances center (i, j), the bounding boxes of instances (top, left, height, width) or the instance segmentation of the image (H, W).

Positions are stored as floats (row first). Bounding boxes are stored as ints (row first), thus right = left + width - 1 The labels of the segmentation mask have to be consecutive. You can make it consecutive using relabel_consecutive.

Additional optional data is also expected like “confidence” or “shape” that respectively defines the confidence for each detection and the shape of the image (H, W).

Any additional meta information on the detections can be also given.

Defines position, bbox, segmentation, and confidence properties. Each of them are either from the data or extrapolated if missing (see _extrapolate_xxx).

data

Detections data.

Type:

Dict[str, toch.Tensor]

length

Number of detections

Type:

int

frame_id

Frame id in the video (-1 if no video) Default: -1

Type:

int

shape

(Tuple[int, int]): Shape of the image (H, W). (Extrapolated if not given)

position

Positions (i, j) of instances (center) inferred from the data Shape: (N, 2), dtype: float32

Type:

torch.Tensor

bbox

Bounding boxes of instances inferred from the data Shape: (N, 4), dtype: int32

Type:

torch.Tensor

segmentation

Segmentation inferred from the data Shape: (H, W), dtype: int32

Type:

torch.Tensor

confidence

Confidence for each instance Shape: (N,), dtype: float32

Type:

torch.Tensor

save(path: str | PathLike) → None

Save detections to a file using torch.save

Parameters:

path (str | os.PathLike) – Output path

static load(path: str | PathLike) → Detections

Load a detections for a given frame using torch.load

Parameters:

path (str | os.PathLike) – Input path

static save_multi_frames_detections(detections_sequence: Collection[Detections], path: str | PathLike) → None

Save detections for a sequence of frames

It will save the detections as:

path/{frame_id}.pt
     ...

Parameters:

• detections_sequence (Collection[Detections]) – Detections for each frame Each detections should have a different frame_id

• path (str | os.PathLike) – Output folder

static load_multi_frames_detections(path: str | PathLike) → Collection[Detections]

Load detections for a sequence of frames

Expect the following file structure:

path/{frame_id}.pt
     ...

Parameters:

path (str | os.PathLike) – Input folder

Returns:

Detections for each frame (sorted by frame id)

Return type:

Collection[Detections]