Skip to main content

visualization.preprocessing

logger_config​

logger​

preprocessing_visualization​

@save_state(model=PreprocessingVisualizationFunctionSchema)
def preprocessing_visualization(config: dict,
                                save_to_file: bool = False,
                                show_figure: bool = True) -> None

visualize_preprocessing_scatter​

def visualize_preprocessing_scatter(
        config: dict,
        session_index: int = 0,
        frames: list = [],
        original_positions_key: str | None = "position",
        cleaned_positions_key: str | None = "position_cleaned_lowconf",
        aligned_positions_key: str | None = "position_egocentric_aligned",
        filtered_positions_key: str | None = "position_processed",
        scaled_positions_key: str | None = "position_scaled",
        save_to_file: bool = False,
        show_figure: bool = True)

Visualize the preprocessing results by plotting the positions of the keypoints in a scatter plot. Each position key parameter can be a string (to include that column) or None (to skip that column).

Parameters

  • config (dict): Configuration dictionary.
  • session_index (int, optional): Index of the session to visualize.
  • frames (list, optional): List of frames to visualize.
  • original_positions_key (str, optional): Key for the original positions.
  • cleaned_positions_key (str, optional): Key for the low confidence cleaned positions.
  • aligned_positions_key (str, optional): Key for the egocentric aligned positions.
  • filtered_positions_key (str, optional): Key for the filtered positions.
  • scaled_positions_key (str, optional): Key for the scaled positions.
  • save_to_file (bool, optional): Whether to save the figure to a file.
  • show_figure (bool, optional): Whether to show the figure.

Returns

  • None

visualize_preprocessing_timeseries​

def visualize_preprocessing_timeseries(
        config: dict,
        session_index: int = 0,
        n_samples: int = 1000,
        sample_offset: int = 0,
        original_positions_key: str | None = "position",
        aligned_positions_key: str | None = "position_egocentric_aligned",
        filtered_positions_key: str | None = "position_processed",
        scaled_positions_key: str | None = "position_scaled",
        keypoints: list | None = None,
        save_to_file: bool = False,
        show_figure: bool = True)

Visualize the preprocessing results by plotting position data in a timeseries plot.

Parameters

  • config (dict): Configuration dictionary.
  • session_index (int, optional): Index of the session to visualize.
  • n_samples (int, optional): Number of samples to plot.
  • sample_offset (int, optional): Starting index for the time series data. Default is 0 (start from beginning).
  • original_positions_key (str | None, optional): Key for the original positions. If None, this position type will be skipped.
  • aligned_positions_key (str | None, optional): Key for the aligned positions. If None, this position type will be skipped.
  • filtered_positions_key (str | None, optional): Key for the filtered positions. If None, this position type will be skipped.
  • scaled_positions_key (str | None, optional): Key for the scaled positions. If None, this position type will be skipped.
  • keypoints (list | None, optional): List of keypoint names to include in the visualization. If None or empty list, all keypoints will be included.
  • save_to_file (bool, optional): Whether to save the figure to a file.
  • show_figure (bool, optional): Whether to show the figure.

Returns

  • None

visualize_preprocessing_cloud​

def visualize_preprocessing_cloud(
        config: dict,
        session_index: int = 0,
        n_samples: int = 1000,
        aligned_positions_key: str | None = "position_egocentric_aligned",
        filtered_positions_key: str | None = "position_processed",
        scaled_positions_key: str | None = "position_scaled",
        keypoints: list | None = None,
        alpha: float = 0.3,
        save_to_file: bool = False,
        show_figure: bool = True)

Visualize the preprocessing results by plotting a cloud of keypoint positions across multiple frames. Only includes aligned, filtered, and scaled positions as these are in comparable coordinate systems.

Parameters

  • config (dict): Configuration dictionary.
  • session_index (int, optional): Index of the session to visualize.
  • n_samples (int, optional): Number of frames to include in the visualization. Frames are randomly sampled.
  • aligned_positions_key (str | None, optional): Key for the egocentric aligned positions. If None, this position type will be skipped.
  • filtered_positions_key (str | None, optional): Key for the filtered positions. If None, this position type will be skipped.
  • scaled_positions_key (str | None, optional): Key for the scaled positions. If None, this position type will be skipped.
  • keypoints (list | None, optional): List of keypoint names to include. If None, all keypoints will be included.
  • alpha (float, optional): Transparency level for the dots (0.0 to 1.0).
  • save_to_file (bool, optional): Whether to save the figure to a file.
  • show_figure (bool, optional): Whether to show the figure.

Returns

  • None