deepof.annotation_utils

Functions and general utilities for supervised pose estimation. See documentation for details.

Functions

climb_wall(arena_type, arena, pos_dict, tol, ...)	Return True if the specified mouse is climbing the wall.
close_double_contact(pos_dframe, left1, ...)	Return a boolean array that's True if the specified body parts are closer than tol.
close_single_contact(pos_dframe, left, ...)	Return a boolean array that's True if the specified body parts are closer than tol.
following_path(distance_dframe, ...[, ...])	Return True if 'follower' is closer than tol to the path that followed has walked over the last specified number of frames.
frame_corners(w, h[, corners])	Return a dictionary with the corner positions of the video frame.
get_hparameters([hparams])	Return the most frequent behaviour in a window of window_size frames.
huddle(X_huddle, huddle_estimator[, animal_id])	Return true when the mouse is huddling a pretrained model.
look_around(speed_dframe, likelihood_dframe, ...)	Return true when the mouse is looking around using simple rules.
max_behaviour(behaviour_dframe[, ...])	Return the most frequent behaviour in a window of window_size frames.
outside_ellipse(x, y, e_center, e_axes, e_angle)	Auxiliar function to climb_wall and sniff_object.
rotate(origin, point, ang)	Auxiliar function to climb_wall and sniff_object.
sniff_object(speed_dframe, arena_type, ...)	Return True if the specified mouse is sniffing an object.
supervised_tagging(coord_object, raw_coords, ...)	Output a dataframe with the registered motives per frame.
tagged_video_output(coordinates, tag_dict[, ...])	Output annotated videos.