Module misty2py_skills.face_recognition

This module implements a face recognition skill that allows Misty to greet people with their chosen name.

Global variables

var TESTING_NAME

The name of the person testing this module. For convenience, the faces commencing with this name are forgotten in at beginning of the skill so they can be learnt again.

var UNKNOWN_LABEL

The label used by Misty's REST API for an unknown face.

var UPDATE_TIME

The minimum amount of time (in seconds) that must pass between Misty greets the same person again.

Functions

def face_recognition(misty: misty2py.robot.Misty) ‑> Dict

Expand source code

def face_recognition(misty: Misty) -> Dict:
    """Misty detects a face, if she knows the person, she greets them by their name, else she prompts them to join a face training session.

    Args:
        misty (Misty): The Misty to perform the skill.

    Returns:
        Dict: The dictionary with `"overall_success"` key (bool) and keys for every action performed (dictionarised Misty2pyResponse).
    """
    cancel_skills(misty)
    set_volume = misty.perform_action(
        "volume_settings", data="low_volume"
    ).parse_to_dict()

    get_faces_known = (
        misty.get_info("faces_known").parse_to_dict().get("rest_response", {})
    )
    known_faces = get_faces_known.get("result")
    if not known_faces is None:
        print("Your misty currently knows these faces: %s." % ", ".join(known_faces))
        purge_testing_faces(misty, known_faces)
    else:
        print("Your Misty currently does not know any faces.")

    start_face_recognition = misty.perform_action(
        "face_recognition_start"
    ).parse_to_dict()

    subscribe_face_recognition = misty.event(
        "subscribe", type="FaceRecognition", name=event_face_rec, event_emitter=ee
    ).parse_to_dict()
    status.set_(status=StatusLabels.MAIN)
    print(subscribe_face_recognition)

    print(">>> Type 'stop' to terminate <<<")
    user_input = ""
    while not (
        user_input in UserResponses.STOP and status.get_("status") == StatusLabels.MAIN
    ):
        user_input = input().lower()
        handle_user_input(misty, user_input)

    unsubscribe_face_recognition = misty.event(
        "unsubscribe", name=event_face_rec
    ).parse_to_dict()
    print(unsubscribe_face_recognition)

    face_recognition_stop = misty.perform_action(
        "face_recognition_stop"
    ).parse_to_dict()

    speak_wrapper(misty, "Bye!")

    return success_of_action_dict(
        set_volume=set_volume,
        get_faces_known=get_faces_known,
        start_face_recognition=start_face_recognition,
        subscribe_face_recognition=subscribe_face_recognition,
        unsubscribe_face_recognition=unsubscribe_face_recognition,
        face_recognition_stop=face_recognition_stop,
    )

Misty detects a face, if she knows the person, she greets them by their name, else she prompts them to join a face training session.

Args

misty : Misty

The Misty to perform the skill.

Returns

Dict

The dictionary with "overall_success" key (bool) and keys for every action performed (dictionarised Misty2pyResponse).

def handle_greeting(misty: misty2py.robot.Misty, user_name: str) ‑> None

Expand source code

def handle_greeting(misty: Misty, user_name: str) -> None:
    """Greets a person."""
    status.set_(status=StatusLabels.GREET)
    utterance = f"Hello, {user_from_face_id(user_name)}!"
    speak_wrapper(misty, utterance)
    status.set_(status=StatusLabels.MAIN)

Greets a person.

def handle_recognition(misty: misty2py.robot.Misty, label: str, det_time: float) ‑> None

Expand source code

def handle_recognition(misty: Misty, label: str, det_time: float) -> None:
    """Handles recognition of a face."""
    status.set_(data=label, time=det_time)
    if label == UNKNOWN_LABEL:
        training_prompt(misty)
    else:
        handle_greeting(misty, label)

Handles recognition of a face.

def handle_user_input(misty: misty2py.robot.Misty, user_input: str) ‑> None

Expand source code

def handle_user_input(misty: Misty, user_input: str) -> None:
    """Handles keyboard inputs."""
    if user_input in UserResponses.YES and status.get_("status") == StatusLabels.PROMPT:
        initialise_training(misty)

    elif (
        status.get_("status") == StatusLabels.INIT
        and not user_input in UserResponses.STOP
    ):
        perform_training(misty, user_input)

    elif (
        status.get_("status") == StatusLabels.INIT and user_input in UserResponses.STOP
    ):
        d = misty.perform_action("face_train_cancel").parse_to_dict()
        print(d)
        status.set_(status=StatusLabels.MAIN)

    elif (
        status.get_("status") == StatusLabels.TALK and user_input in UserResponses.STOP
    ):
        d = misty.perform_action("speak_stop").parse_to_dict()
        print(d)
        status.set_(status=StatusLabels.MAIN)

    elif status.get_("status") == StatusLabels.PROMPT:
        print("Training not initialised.")
        status.set_(status=StatusLabels.MAIN)

Handles keyboard inputs.

def initialise_training(misty: misty2py.robot.Misty)

Expand source code

def initialise_training(misty: Misty):
    """Initialises the face training session, including prompting the user to enter their name."""
    status.set_(status=StatusLabels.INIT)
    speak_wrapper(
        misty,
        "<p>How should I call you?</p><p>Please enter your name in the terminal.</p>",
    )
    print("Enter your name (the first name suffices)")

Initialises the face training session, including prompting the user to enter their name.

def listener(data: Dict) ‑> None

Expand source code

@ee.on(event_face_train)
def listener(data: Dict) -> None:
    """Handles face training events and terminates the training session if the data states the training is complete.

    Args:
        data (Dict): The data received from Misty's WebSocket API along with the face training event.
    """
    if data.get("message") == "Face training embedding phase complete.":
        speak_wrapper(misty_glob, "Thank you, the training is complete now.")
        status.set_(status=StatusLabels.MAIN)
        d = misty_glob.event("unsubscribe", name=event_face_train).parse_to_dict()
        print(d)

Handles face training events and terminates the training session if the data states the training is complete.

Args

data : Dict

The data received from Misty's WebSocket API along with the face training event.

def perform_training(misty: misty2py.robot.Misty, name: str) ‑> None

Expand source code

def perform_training(misty: Misty, name: str) -> None:
    """Performs the training.

    Args:
        misty (Misty): The Misty to learn the user's face.
        name (str): The name of the user.
    """
    status.set_(status=StatusLabels.TRAIN)
    d = misty.get_info("faces_known").parse_to_dict().get("rest_response", {})
    new_name = name
    if not d.get("result") is None:
        while new_name in d.get("result"):
            new_name = name + "_" + get_random_string(3)
        if new_name != name:
            print(f"The name {name} is already in use, using {new_name} instead.")
    if new_name == "":
        new_name = get_random_string(6)
        print(f"The name {name} is invalid, using {new_name} instead.")

    d = misty.perform_action(
        "face_train_start", data={"FaceId": new_name}
    ).parse_to_dict()
    print(d)
    speak_wrapper(misty, "The training has commenced, please do not look away now.")
    misty.event(
        "subscribe", type="FaceTraining", name=event_face_train, event_emitter=ee
    )

Performs the training.

Args

misty : Misty

The Misty to learn the user's face.

name : str

The name of the user.

def purge_testing_faces(misty: misty2py.robot.Misty, known_faces: List) ‑> None

Expand source code

def purge_testing_faces(misty: Misty, known_faces: List) -> None:
    """Forgets the faces of users that start with `TESTING NAME`."""
    for face in known_faces:
        if face.startswith(TESTING_NAME):
            d = (
                misty.perform_action("face_delete", data={"FaceId": face})
                .parse_to_dict()
                .get("rest_response", {})
            )
            if d.get("success"):
                print("Successfully forgot the face of %s." % face)
            else:
                print("Failed to forget the face of %s." % face)

Forgets the faces of users that start with TESTING NAME.

def speak_wrapper(misty: misty2py.robot.Misty, utterance: str) ‑> None

Expand source code

def speak_wrapper(misty: Misty, utterance: str) -> None:
    """Changes status to `StatusLabels.TALK` while Misty is talking."""
    prev_stat = status.get_("status")
    status.set_(status=StatusLabels.TALK)
    print(speak(misty, utterance))
    status.set_(status=prev_stat)

Changes status to StatusLabels.TALK while Misty is talking.

def training_prompt(misty: misty2py.robot.Misty)

Expand source code

def training_prompt(misty: Misty):
    """Misty asks whether to begin a face training session."""
    status.set_(status=StatusLabels.PROMPT)
    speak_wrapper(
        misty,
        "Hello! I do not know you yet, do you want to begin the face training session?",
    )
    print("An unknown face detected.\nDo you want to start training (yes/no)? [no]")

Misty asks whether to begin a face training session.

def user_from_face_id(face_id: str) ‑> str

Expand source code

def user_from_face_id(face_id: str) -> str:
    """Returns the name from a face ID."""
    return face_id.split("_")[0].capitalize()

Returns the name from a face ID.

Classes

class StatusLabels (value, names=None, *, module=None, qualname=None, type=None, start=1)

Expand source code

class StatusLabels(Enum):
    """Represents states in which the skill can be."""

    MAIN = "running_main"
    """The main state of waiting for a face recognition event."""
    PROMPT = "running_train_prompt"
    """The state of running the training prompt and waiting for the user's response."""
    TRAIN = "running_training"
    """The face training state."""
    INIT = "running_training_initialisation"
    """The state of the initialisation of face training."""
    GREET = "running_greeting"
    """The state of greeting a person."""
    TALK = "talking"
    """The state of talking."""

Represents states in which the skill can be.

Ancestors

• enum.Enum

Class variables

var GREET

The state of greeting a person.

var INIT

The state of the initialisation of face training.

var MAIN

The main state of waiting for a face recognition event.

var PROMPT

The state of running the training prompt and waiting for the user's response.

var TALK

The state of talking.

var TRAIN

The face training state.

class UserResponses

Expand source code

class UserResponses:
    """Represents responses that a user can give to different prompts used in this skill."""

    YES = frozenset(["yes", "y"])
    NO = frozenset(["no", "n"])
    STOP = frozenset(["stop", "s", "terminate"])

Represents responses that a user can give to different prompts used in this skill.

Class variables

var NO

var STOP

var YES