Function analyze_images

deeplabcut/compat.py:1139–1312

Analyzes images with a DeepLabCut model and stores the output in an H5 file. This method is only implemented for PyTorch models. The labels are stored as Pandas DataFrame, which contains the name of the network, body part name, (x, y) label position in pixels, and the likelihood for ea

(
    config: str | Path,
    images: str | Path | list[str] | list[Path],
    frame_type: str | None = None,
    destfolder: str | Path | None = None,
    shuffle: int = 1,
    trainingsetindex: int = 0,
    max_individuals: int | None = None,
    device: str | None = None,
    snapshot_index: int | None = None,
    detector_snapshot_index: int | None = None,
    save_as_csv: bool = False,
    modelprefix: str = "",
    plotting: bool | str = False,
    pcutoff: float | None = None,
    bbox_pcutoff: float | None = None,
    plot_skeleton: bool = False,
    **torch_kwargs,
)

def analyze_images(
    config: str | Path,
    images: str | Path | list[str] | list[Path],
    frame_type: str | None = None,
    destfolder: str | Path | None = None,
    shuffle: int = 1,
    trainingsetindex: int = 0,
    max_individuals: int | None = None,
    device: str | None = None,
    snapshot_index: int | None = None,
    detector_snapshot_index: int | None = None,
    save_as_csv: bool = False,
    modelprefix: str = "",
    plotting: bool | str = False,
    pcutoff: float | None = None,
    bbox_pcutoff: float | None = None,
    plot_skeleton: bool = False,
    **torch_kwargs,
) -> dict[str, dict[str, np.ndarray | np.ndarray]]:
    """Analyzes images with a DeepLabCut model and stores the output in an H5 file.

    This method is only implemented for PyTorch models.

    The labels are stored as Pandas DataFrame, which contains the name of the network,
    body part name, (x, y) label position in pixels, and the likelihood for each frame
    per body part.

    Parameters
    ----------
    config : str, Path
        Full path of the project's config.yaml file.

    images: str, Path, list[str], list[Path]
        The image(s) to run inference on. Can be the path to an image, the path
        to a directory containing images, or a list of image paths or directories
        containing images.

    frame_type: string, optional
        Filters the images to analyze to only the ones with the given suffix (e.g.
        setting `frame_type`=".png" will only analyze ".png" images). The default
        behavior analyzes all ".jpg", ".jpeg" and ".png" images.

    destfolder: str, Path, optional
        The directory where the predictions will be stored. If None, the predictions
        will be stored in the same directory as the first image given in the `images`
        argument (if it's a directory, that directory will be used; if it's an image,
        the directory containing the image will be used).

    shuffle: int, optional
        An integer specifying the shuffle with which to run image analysis.

    trainingsetindex: int, optional
        Integer specifying which TrainingsetFraction to use. By default, the first one
        is used (note that TrainingFraction is a list in config.yaml).

    max_individuals: int, optional
        The maximum number of individuals to detect in each image. Set to the number of
        individuals in the project if None.