Reshape inputs to expected shape.

Check if the onnx file and prototxt file exists, and if necessary, download the files to the given path. Parameters ---------- w

Copies this config into a new config object, making the changes given by new_config_dict.

Get ailia default argument parser Parameters ---------- description : str default_input : str default input data (image

Default check or update configurations should be placed here Parameters ---------- parser : ArgumentParser() Returns -------

Propagate the state distribution to the current time step using a Kalman filter prediction step. Parameters ----------

Get savepath NOTE: we may have better option... TODO: args.save_dir & args.save_path ? Parameters ---------- arg_path : str

Get label with longest duration Parameters ---------- support : Segment or Timeline, optional Find label with lon

Copies new_config_dict into this config object. Note: new_config_dict can also be a config object.

Estimates the head pose given an image. Parameters ---------- img: NumPy array The image in BGR channels

:param detector: ailia.Detector, or list of ailia.DetectorObject :param img: ndarray data of image :param category: list of category_name

list of all keys

Get cv2.VideoCapture * TODO: maybe get capture & writer at the same time? * then, you can use capture frame size directory Pa

Flattens any np.array of column vectors into 1D arrays. Basically, this makes data readable for humans if you are just inspecting via

Instantiate a PreTrainedBertModel from a pre-trained model file. Download and cache the pre-trained model file if needed.

Normalize image Parameters ---------- image: numpy array The image you want to normalize normalize_type: string

get cv2.VideoWriter Parameters ---------- save_path : str height : int width : int fps : int rgb : bool, default is True

Add a new hypothesis to the list.

Loads the image of the given path, performs the necessary preprocessing, and returns it. Parameters ---------- image_path: strin

Check if the file exists, and if necessary, download the files to the given path. Parameters ---------- file_path: string

save the current state of the Kalman filter

Given a matrix of character probabilities, returns the decoder's best guess of the transcription Arguments: prob

Advances the state vector and returns the predicted bounding box estimate.

Remove a segment (in place) Parameters ---------- segment : Segment Segment that is being removed Return

Get index of (existing) segment Parameters ---------- segment : Segment Segment that is being looked for.

Search new tokens for running hypotheses and encoded speech x. Args: running_hyps (List[Hypothesis]): Running hypotheses on beam

Fast version of self(file).crop(segment, **kwargs) Parameters ---------- file : AudioFile Audio file. seg

Generates a pretty printed NumPy array with an assignment. Optionally transposes column vectors so they are drawn on one line. Strictly speak

Get sorted list of labels Returns ------- labels : list Sorted list of labels

Get a copy of the segment Returns ------- copy : Segment Copy of the segment.

Load pretrained pipeline Parameters ---------- checkpoint_path : Path or str Path to pipeline checkpoint, or a re

Perform Kalman filter measurement update step and update the feature cache. Parameters ---------- kf : kalman_filter.

Return an empty copy Returns ------- empty : Timeline Empty timeline using the same 'uri' attribute.

Get timeline made of all annotated segments Parameters ---------- copy : bool, optional Defaults (True) to return

Estimates a box 9-dof parameters from the given vertices. Directly computes the scale of the box, then solves for orientation and tra

Iterate over tracks (in chronological order) Parameters ---------- yield_label : bool, optional When True, yield

Draws the head pose(s) on the image. (Person POV) The axes correspond to x (blue) oriented positively to the left, y (green) oriented

Add a segment (in place) Parameters ---------- segment : Segment Segment that is being added Returns

Converts a sequence of ids in wordpiece tokens using the vocab.

Predict the gaze given an image. Parameters ---------- img : NumPy array The image in BGR channels. gazes