Abstract: A method and an electronic apparatus for performing unified segmentation of media content are provided. The method includes: determining a guidance map for an input frame based on a salient object from a past frame output mask and user-interacted objects in the media, operating in either salient mode or selective mode. The input frame of the media is cropped based on the guidance map and the salient ROIs of the salient object. A weighted grayscale image of the cropped frame is generated from the past frame output mask. A fused spatio-color mesh grid representation of the cropped frame in YUV format is determined. The cropped image frame, along with the weighted grayscale image and the fused spatio-color mesh grid representation, is input into a segmentation model. The segmentation model generates either a salient object segmentation or a user-interacted object segmentation for the media.

Abstract: Devices, methods, and systems for aircraft detection, tracking, and docking using three-dimensional sensor data are described herein. One method includes capturing, by a light detection and ranging (LIDAR) sensor, LIDAR sensor data of an aircraft while the aircraft is approaching a docking area of an airport, receiving, from the LIDAR sensor, the sensor data of the aircraft while the aircraft is approaching a docking area, retrieving, by a computing device, a simulated two-dimensional image of the aircraft generated based on simulated data of the aircraft, tracking, by the computing device, a position of the aircraft while the aircraft is approaching the docking area using the LIDAR sensor data and the simulated two-dimensional image, and displaying an image including the position of the aircraft while the aircraft is approaching the docking area to provide navigation guidance to a pilot.

Abstract: A method for segmenting objects in a scene by an electronic device is provided. The method includes inputting at least one input frame of the scene into a pre-trained neural network model, the scene including a plurality of objects; determining a position and a shape of each object of the plurality of objects in the scene using the pre-trained neural network model; determining an array of coefficients for pixels associated with each object of the plurality of objects in the scene using the pre-trained neural network model; and generating a segment mask for each object of the plurality of objects based on the position, the shape, and the array of coefficients for each object of the plurality of objects in the scene.

Abstract: A method of performing saliency segmentation for a preview image frame, including: receiving the preview image frame from an imaging unit, generating a plurality of saliency boxes including one or more salient subjects, for each of a plurality of subjects in the preview image frame, selecting, from among the plurality of saliency boxes, a set of saliency boxes including a first set of salient subjects based on a ranking of each saliency box from among the plurality of saliency boxes, and extracting one or more salient images along with boundary information corresponding to each salient subject from among the first set of salient subjects.

Abstract: A device and method for performing image correction may include: receiving an input image to be corrected; generating a plurality of template images based on the input image, wherein each template image from among the plurality of template images is generated based on at least one of an image feature transformation; providing the plurality of template images to an artificial intelligence (AI) module; assigning, using the AI module, weights from among zero and non-zero magnitudes to each template image based on an extent and a type of correction to be applied to each template image; and applying the assigned weights to the plurality of template images to obtain a corrected output image.

Abstract: A vison-based landing system comprises a processor onboard an aerial vehicle, an onboard vision sensor, onboard aiding sensors, and a data storage unit. The processor hosts an adaptive feature extraction module operative to perform a method that comprises capturing an image of a landing area having a plurality of edge features; calculating an estimated slope of an expected edge feature; calculating an expected gradient direction of the expected edge feature; selecting a horizontal basis kernel for the expected edge feature along an expected horizontal gradient; selecting a vertical basis kernel for the expected edge feature along an expected vertical gradient; calculating a combined convolution kernel for the expected edge feature based on the horizontal and vertical basis kernels, the estimated slope, and the expected gradient direction; and performing a convolution operation on the image using the combined convolution kernel to obtain an edge feature image of the landing area.

Abstract: Systems and methods are provided that may facilitate the irrigation of an anatomic space. The irrigation system can include a tube manifold, a fluid reservoir, a waste fluid trap and an emulsification system. The emulsification system can be arranged at a distal end portion of the tube manifold and can be configured to be inserted into an anatomic space and emulsify organic material located within the anatomic space.

Abstract: Disclosed are methods, systems, and non-transitory computer-readable medium for managing energy use in a vehicle. For instance, the method may include receiving forecasted data from a first external source, receiving real-time data corresponding to at least one weather parameter at a first location at a first time, and continuously determining whether to perform an adjustment to a control parameter of the vehicle by using a machine learning model that is based on the forecasted data for the at least one weather parameter, the real-time data for the at least one weather parameter, a battery condition of the vehicle, and/or an estimated amount of energy consumed by traveling along a first navigation path.

Abstract: In a method of encoding of temporal information for stable video instance segmentation and video object detection, a neural network analyzes an input frame of a video to output a prediction template. The prediction template includes either segmentation masks of objects in the input frame or bounding boxes surrounding objects in the input frame. The prediction template is then colour coded by a template generator. The colour coded template, along with a frame subsequent to the input frame, is supplied to a template encoder such that temporal information from the input frame is encoded into the output of the temporal encoder.

Abstract: A vison-based landing system comprises a processor onboard an aerial vehicle, an onboard vision sensor, onboard aiding sensors, and a data storage unit. The processor hosts an adaptive feature extraction module operative to perform a method that comprises capturing an image of a landing area having a plurality of edge features; calculating an estimated slope of an expected edge feature; calculating an expected gradient direction of the expected edge feature; selecting a horizontal basis kernel for the expected edge feature along an expected horizontal gradient; selecting a vertical basis kernel for the expected edge feature along an expected vertical gradient; calculating a combined convolution kernel for the expected edge feature based on the horizontal and vertical basis kernels, the estimated slope, and the expected gradient direction; and performing a convolution operation on the image using the combined convolution kernel to obtain an edge feature image of the landing area.

Abstract: Methods and systems are provided for slewing a light beam axis directly between points on the ground. One method involves determining a first position associated with a beam axis of a lighting arrangement in a Cartesian reference frame based on an initial orientation of the lighting arrangement in a spherical reference frame, determining an adjustment for the lighting arrangement in the Cartesian reference frame in response to a user input, determining an updated position for the beam axis in the Cartesian reference frame based on the first position and the adjustment in the Cartesian reference frame, transforming the updated position for the beam axis in the Cartesian reference frame to an updated orientation of the lighting arrangement in the spherical reference frame, and concurrently commanding actuators associated with the lighting arrangement to slew the lighting arrangement from the initial orientation to the updated orientation in the spherical reference frame.

Abstract: A method for controlling an orientation of a searchlight on a vehicle is provided. The method comprises: obtaining a target range to a point of interest (POI) at a target; determining a searchlight position and attitude; calculating a three dimensional position at the POI using searchlight azimuth and tilt actuator angles, the target range, and the searchlight position and attitude, while a searchlight light head continues to point at the 3D target position despite changes in movement or orientation of the vehicle; calculating a desired searchlight orientation including azimuth angle and tilt angle to point the light head at the target through inverting a kinematic relationship between the vehicle and the searchlight; calculating compensatory actuator angles for a plurality of searchlight actuators to achieve the desired searchlight orientation based on error measurements calculated from a current orientation; and commanding the plurality of searchlight actuators to the compensatory actuator angles.

Abstract: A method for segmenting objects in a scene by an electronic device is provided. The method includes inputting at least one input frame of the scene into a pre-trained neural network model, the scene including a plurality of objects; determining a position and a shape of each object of the plurality of objects in the scene using the pre-trained neural network model; determining an array of coefficients for pixels associated with each object of the plurality of objects in the scene using the pre-trained neural network model; and generating a segment mask for each object of the plurality of objects based on the position, the shape, and the array of coefficients for each object of the plurality of objects in the scene.

Abstract: A gimbal assembly is provided that allows for more than a single rotation about a rotational axis in both a first rotational direction and a second rotational direction.

Abstract: Systems and Embodiments for a structured light navigation aid are provided. In certain embodiments, a system includes a structured light emitter that emits structured light towards a surface, wherein the structured light emitter can movably change a direction of emitted structured light. Additionally, the system includes a structured light receiver that receives light from an environment, wherein the structured light receiver identifies reflected structured light from the surface in the received light and calculates navigation information from the reflected structured light. Further, the system includes one or more processors that control the structured light as one of a pre-programmed pattern or a pattern determined during operation of the structured light emitter.

Abstract: A method of tracking an image feature is described. The method comprises acquiring an image with a camera, and determining, using processing circuitry, a bounding area in the image, the bounding area surrounding a feature in the image. The method further comprises determining, using processing circuitry, a rotation axis and a rotation angle based on a first focal length of the camera and a position of the bounding area relative to a center of the image. The method further comprises determining, using processing circuitry, at least one of a pan angle, a roll angle, and a tilt angle for the camera at which the bounding area is centered in the image. The method further comprises adjusting, using a gimbal, an orientation of the camera based on the at least one of the pan angle, the roll angle, and the tilt angle.

Abstract: Disclosed are methods, systems, and non-transitory computer-readable medium for managing energy use in a vehicle. For instance, the method may include receiving forecasted data from a first external source, receiving real-time data corresponding to at least one weather parameter at a first location at a first time, and continuously determining whether to perform an adjustment to a control parameter of the vehicle by using a machine learning model that is based on the forecasted data for the at least one weather parameter, the real-time data for the at least one weather parameter, a battery condition of the vehicle, and/or an estimated amount of energy consumed by traveling along a first navigation path.

Abstract: A gimbal assembly is provided that allows for more than a single rotation about a rotational axis in both a first rotational direction and a second rotational direction.

Abstract: A method of tracking an image feature is described. The method comprises acquiring an image with a camera, and determining, using processing circuitry, a bounding area in the image, the bounding area surrounding a feature in the image. The method further comprises determining, using processing circuitry, a rotation axis and a rotation angle based on a first focal length of the camera and a position of the bounding area relative to a center of the image. The method further comprises determining, using processing circuitry, at least one of a pan angle, a roll angle, and a tilt angle for the camera at which the bounding area is centered in the image. The method further comprises adjusting, using a gimbal, an orientation of the camera based on the at least one of the pan angle, the roll angle, and the tilt angle.

Abstract: An active human-machine interface feedback system includes a user interface, a pitch angle sensor, a roll angle sensor, a spherical motor, and a control circuit. The user interface adapted to receive user input and is configured, upon receipt of the user input, to move, about one or both of a pitch axis and a roll axis, to a user interface position. The pitch angle sensor is configured to sense the pitch angle component of the user interface position. The roll angle sensor is configured to sense the roll angle component of the user interface position. The spherical motor is coupled to the user interface and is symmetrically disposed about the origin. The control circuit determines a polar angle (?) of the user interface relative to the origin, determine an azimuthal angle (?) of the user interface relative to the origin, and supply current to the first, second, and third coils.