Abstract: Disclosed is a method of tracking an object and apparatuses performing the same, the method including generating a plurality of candidate particles corresponding to an object in a video frame to track the object, and tracking the object by generating optimal particles corresponding to a position of the object in the video frame based on similarities of the plurality of candidate particles with respect to the object.

Abstract: A method of object surface matching includes identifying an object in-flight in an image; identifying a feature on the object that is in a first spatial position; comparing the feature with set of template images; identifying a first template image in the set of template images that matches the feature on the object that is in the first spatial position; determining first coordinates for the first spatial position based on the first template image; identifying a second image of the object that includes the feature on the object that is in a second spatial position; identifying a second template image in the set of template images that matches the feature on the object that is in the second spatial position; determining second coordinates for the second spatial position based on the second template image; and generating a spin value for the object based on the first and second coordinates.

Abstract: A computer-implemented method and system for selecting one or more regions of interest (ROIs) in an image. The method comprises: identifying one or more objects of interest that have been segmented from the image; identifying predefined landmarks of the objects; determining reference morphometrics pertaining to the objects by performing morphometrics on the objects by reference to the landmarks; selecting one or more ROIs from the objects according to the reference morphometrics, comprises identifying the location of the ROIs relative to the reference morphometrics; and outputting the selected one or more ROIs.

Abstract: The present invention falls within the technical field of robots, and particularly relates to a robot monitoring system. The robot monitoring system based on human body information comprises an image collection unit for collecting an image; a human body detection unit for determining whether human body information matching a training sample exists in the image; a target acquisition unit for acquiring position information and size information about the human body information in the image; a target tracking unit for acquiring a current motion region according to a frame difference between the image and a predetermined image frame and acquiring a moving target from the motion region; and an adjustment unit for adjusting the orientation of the image collection unit such that the moving target is located in the centre of the image.

Abstract: Image processing of an image is used to determine the opportunity to view an object. Rather than relying on simple numbers passing an object, the opportunity to view the object is weighted based on attention, which is derived from other objects competing for attention. For the processor to more accurately determine opportunity to view as compared to using geometric information alone, a machine-learned network is used. To deal with changes in obstructions, another machine-learned network may extract obstructions from camera images. Trace data is used to allow for daily variation in base counts of viewers, allowing greater temporal resolution and determination based on information more recently acquired than counts.

Abstract: A method and system are provided for ensuring creation of authentic online profiles. The method includes providing an option to a user for enabling capturing of digital media depicting the user's appearance. Further, digital media using a camera associated with a user device is captured, if the user chooses to allow the capturing of digital media. Furthermore, at least one instruction is communicated to the user for performing one or more actions. Subsequently, the captured digital media is transmitted for verification, without allowing the user to modify the captured digital media. Thereafter, the captured digital media is verified by analyzing the captured digital media for approving or rejecting the captured digital media.

Abstract: A method and system for monitoring the motion of one or both eyes, includes capturing a sequence of overlapping images of a subject's face including an eye and the corresponding non-eye region; identifying a plurality of keypoints in each image; mapping corresponding keypoints in two or more images of the sequence; assigning the keypoints to the eye and to the corresponding non-eye region; calculating individual velocities of the corresponding keypoints in the eye and the corresponding non-eye region to obtain a distribution of velocities; extracting at least one velocity measured for the eye and at least one velocity measured for the corresponding non-eye region; calculating the eye-in-head velocity for the eye based upon the measured velocity for the eye and the measured velocity for the corresponding non-eye region; and calculating the eye-in-head position based upon the eye-in-head velocity.

Abstract: A controller for executing a task based on probabilistic image-based landmark localization, uses a neural network, which is trained to process images of objects of a type having a structured set of landmarks to produce a parametric probability distribution defined by values of parameters for a location of each landmark in each processed image. The controller submits the set of input images to the neural network to produce the values of the parameters that define the parametric probability distribution over the location of each landmark in the structured set of landmarks of each input image. Further, the controller determines, for each input image, a global landmark uncertainty for the image based on the parametric probability distributions of landmarks in the input image and executes the task based on the parametric probability distributions of landmarks in each input image and the global landmark uncertainty of each input image.

Abstract: There is a desire to accurately learn a detection model. Provided is a computer-implemented method including acquiring an input image; acquiring an annotated image designating a region of interest in the input image; inputting the input image to a detection model that generates an output image showing a target region from the input image; calculating an error between the output image and the annotated image, using a loss function that weights an error inside the region of interest more heavily than an error outside the region of interest; and updating the detection model in a manner to reduce the error.

Abstract: An atmospheric precipitation rate computation system, comprising an electronic digital image/video capture apparatus to capture digital images/videos of an environment in which an atmospheric precipitation is taking place; an electronic digital image processing apparatus connected to the electronic digital image/video capture apparatus to receive therefrom and process captured digital images/videos to compute atmospheric precipitation rates of atmospheric precipitations which are taking place in the environments shown in the received digital images/videos; and an electronic display device connected to the electronic digital image processing apparatus to receive therefrom and display the computed atmospheric precipitation rate.

Abstract: A system and method for simultaneously presenting Doppler signals of a Multi-Gated Doppler (MGD) signal corresponding to different anatomical structures is provided. The method includes receiving an MGD signal having a plurality of Doppler signals. The method includes analyzing the MGD signal to select multiple gates, each of the gates corresponding with a Doppler signal, and each of the selected gates associated with a different anatomical structure. The method includes selecting a set of parameters for each of the selected gates and applying each selected set of parameters for each of the selected gates. The selected set of parameters for each of the selected gates may be one or both of image acquisition parameters or display processing parameters. The method includes simultaneously presenting the Doppler signal for each of the selected gates at a display system after the each selected set of parameters is applied.

Abstract: An image processing device includes a transfer function input unit to which each of transfer functions of a plurality of imaging systems is input, a calculation unit that calculates a target resolution value which is a target value of recovery processing that recovers a plurality of captured images to be output from each of the plurality of the imaging systems based on the input transfer functions and a predetermined criterion, a recovery-filter generation unit that generates a recovery-filter used for the recovery processing with respect to each of the plurality of imaging systems based on the transfer functions of the plurality of imaging systems and a target resolution value, and a recovered image generation unit that performs the recovery processing with respect to the captured images acquired from the plurality of imaging systems by using the recovery-filter generated for each of the plurality of imaging systems to generate recovered images.

Abstract: A vehicle control system includes: a recognizer that is configured to recognize a surroundings status of a vehicle; and a driving controller that is configured to control at least steering of the vehicle on the basis of the surroundings status recognized by the recognizer, wherein the driving controller is configured to cause the vehicle to run on the basis of a first reference line deviating to one side of a road in a case in which a target obstacle is not recognized in an advancement direction of the vehicle by the recognizer and, is configured to cause the vehicle to avoid the target obstacle after causing the vehicle to run on the basis of a second reference line disposed on an other side of the road from the first reference line, in a case in which the target obstacle is recognized by the recognizer.

Abstract: A computing system determines accuracy of sport-related information extracted from a time sequence of digital video frames that represent a sport event, the extracted sport-related information including an attribute that changes over the time sequence. The computing system (a) detects, based on the extracted sport-related information, a pattern of change of the attribute over the time sequence and (b) makes a determination of whether the detected pattern is an expected pattern of change associated with the sport event. If the determination is that the detected pattern is the expected pattern, then, responsive to making the determination, the computing system takes a first action that corresponds to the sport-related information being accurate. Whereas, if the determination is that the detected pattern is not the expected pattern, then, responsive to making the determination, the computing system takes a second action that corresponds to the sport-related information being inaccurate.

Abstract: A computer-implemented method and system are provided for training a model for New Class Categorization (NCC) of a test image. The method includes decoupling, by a hardware processor, a feature extraction part from a classifier part of a deep classification model by reparametrizing learnable weight variables of the classifier part as a combination of learnable variables of the feature extraction part and of a classification weight generator of the classifier part. The method further includes training, by the hardware processor, the deep classification model to obtain a trained deep classification model by (i) learning the feature extraction part as a multiclass classification task, and (ii) episodically training the classifier part by learning a classification weight generator which outputs classification weights given a training image.

Abstract: A method is provided for visual inspection. The method includes learning, by a processor, group disentangled visual feature embedding vectors of input images. The input images include defective objects and defect-free objects. The method further includes generating, by the processor using a weight generation network, classification weights from visual features and semantic descriptions. Both the visual features and the semantic descriptions are for predicting defective and defect-free labels. The method also includes calculating, by the processor, a cosine similarity score between the classification weights and the group disentangled visual feature embedding vectors. The method additionally includes episodically training, by the processor, the weight generation network on the input images to update parameters of the weight generation network.

Abstract: Systems and methods for contextualizing automatic image segmentation and/or regression including defining an hourglass neural network model, which includes defining an encoder configured to generate compression layers, including a bottleneck layer, and defining a decoder including a contextual convolution operation and configured to generate one or more reconstruction layers. The contextual convolution operation includes establishing, for each of the one or more reconstruction layers, a skip connection between the reconstruction layer and prior layer(s) of different spatial dimension, e.g., between the bottleneck layer and the reconstruction layer and/or between reconstruction layers.

Abstract: A computer includes a processor and a memory storing instructions executable by the processor to receive data from a sensor specifying a plurality of points, the points including a plurality of first points that describe an object; define a boundary surrounding the first points while minimizing a volume of space that is both (i) contained by the boundary and (ii) identified as unoccupied; and actuate a component with respect to a vehicle based on the boundary.

Abstract: Methods and systems are provided for providing directional assistance to guide a user to position a camera for centering a person's face within the camera's field of view. A neural network system is trained to determine the position of the user's face relative to the center of the field of view as captured by an input image. The neural network system is trained using training input images that are generated by cropping different regions of initial training images. Each initial image is used to create a plurality of different training input images, and directional assistance labels used to train the network may be assigned to each training input image based on how the image is cropped. Once trained, the neural network system determines a position of the user's face, and automatically provides a non-visual prompt indicating how to center the face within the field of view.

Abstract: Systems and methods for estimating velocity of an autonomous vehicle and state information of a surrounding vehicle are provided. In some aspects, the system includes a memory that stores instructions for executing processes for estimating velocity of an autonomous vehicle and state information of the surrounding vehicle and a processor configured to execute the instructions. In various aspects, the processes include: receiving image data from an image capturing device; performing a ground plane estimation by predicting a depth of points on a road surface based on an estimated pixel-level depth; determining a three-dimensional (3D) bounding box of the surrounding vehicle; determining the state information of the surrounding vehicle based on the ground plane estimation and the 3D bounding box; and determining the velocity of the autonomous vehicle based on an immovable object relative to the autonomous vehicle.