Abstract: Provided are systems, methods, and devices for generating digital and/or cryptographic assets. An initial state of an environment is acquired using sensors that includes a state of each sensor, a region of interest including a 3D body, and a state of light sources. The asset is associated with the 3D body. A plurality of boundary conditions associated with a workflow for capturing the asset is determined. A visualization of a set of boundary conditions is displayed on a display that includes a plurality of visual cues including first and second visual cues. Each respective visual cue provides a visual indication of a state of a corresponding boundary condition in the set of boundary conditions. At least one visual cue is updated when each boundary condition in the set of boundary conditions is satisfied. When satisfied, the workflow at the computer-enabled imaging device is executed, thereby capturing the asset.

Abstract: A loading lockout system which takes an image of a vessel, recognizes the receiving opening of the vessel, and prevents the loading of material into the vessel until it determines that the discharge footprint of the material storage device is within the receiving opening of the vessel.

Abstract: An estimation result determination unit determines, for each of the objects, a correct estimation result or one of a plurality of false estimation types, which indicate types of false estimation results using ground truth data that corresponds to a video image and output data indicating the result of the estimation made on the video image by the algorithm. The evaluation value calculation unit adds false estimation coefficients that correspond to the plurality of respective false estimation types and are provided so as to become higher in accordance with a degree of impact of the false estimation type for a number of objects that correspond to the false estimation type and thus calculates an evaluation value of the algorithm based on the total value of the added values of the false estimation coefficients obtained for each of the plurality of false estimation types.

Abstract: Systems, methods, and computer-readable storage devices are disclosed for improving markerless motion analysis. One method including: receiving position data of joint centers of a body in motion captured by at least one camera; enhancing, using model equations, three-dimensional (3D) angular kinematic data of the position data of the joint centers of the body, wherein the enhanced 3D angular kinematic data includes increased measurement accuracy of the position data of the joint centers of the body; and providing the enhanced 3d angular kinematic data for display to evaluate motion performance.

Abstract: Metadata is generated for a digital video project in a manner that facilitates mixing video with other content to render a composite video file. The metadata may include global information for the project, per-frame information that acts as a representation of continuous values throughout a video; and key-frame information which is a representation of discrete events throughout the video. The per-frame information may include, for example, point-wise location of specific objects and/or polygons/rectangles of tracked objects. The key-frame information may include information regarding scene changes, information regarding changes in emotion of a person and/or information regarding changes in pose of a person.

Abstract: The present disclosure provides a system and method for improving image resolution of a three-dimensional (3-D) refractive index microscope based on an artificial intelligence (AI) technology. The present disclosure provides a technology for converting a low-resolution 3-D refractive index microscope image into a high-resolution 3-D refractive index image without physical machine conversion and re-photographing based on AI. That is, the present disclosure applies an AI technology, such as deep learning, in order to train an AI model with a physical correlation between a low-resolution 3-D refractive index microscope image and a high-resolution 3-D refractive index image of various samples, such as a cell and a tissue, and convert a low resolution image into a high resolution image without a change in a physical microscope based on the training. Furthermore, for the training of the AI model, the present disclosure uses physical characteristics of a refractive index image.

Abstract: A method includes receiving video data that includes a series of frames of image data. Here, the video data is representative of an actor performing an activity. The method also includes processing the video data to generate a spatial input stream including a series of spatial images representative of spatial features of the actor performing the activity, a temporal input stream representative of motion of the actor performing the activity, and a pose input stream including a series of images representative of a pose of the actor performing the activity. Using at least one neural network, the method also includes processing the temporal input stream, the spatial input stream, and the pose input stream. The method also includes classifying, by the at least one neural network, the activity based on the temporal input stream, the spatial input stream, and the pose input stream.

Abstract: A method for determining a state of a cell, the cell being placed in a sample in contact with a culture medium, includes illuminating the sample with a light source and acquiring an image of the sample with an image sensor, the image sensor lying in a detection plane. From the acquired image, a position of the cell in a plane parallel to the detection plane is located and a refractive index or a relative refractive index of the cell is estimated, the relative refractive index corresponding to a refractive index of the cell relative to the refractive index of the culture medium. From the estimation, an index of interest of the cell is determined. From the index, a state of the cell among predetermined states is classified, the predetermined states including at least one apoptosis state and one living state.

Abstract: A method for detecting wafer cleaning anomalies includes: capturing a wafer cleaning video in real time through each of a plurality of cameras of cleaning machines, each camera corresponds to a respective cleaning chamber of one of the cleaning machines, and each cleaning chamber contains a nozzle; performing image processing on each frame of image contained in the wafer cleaning video to obtain characteristics of contact between a cleaning water column dispensed from the nozzle and a wafer in the image, and determining through the characteristics of contact whether the nozzle has an anomaly; and when a target nozzle having the anomaly is detected, determining anomaly positioning information of the target nozzle, and performing anomaly early-warning by using the anomaly positioning information.

Abstract: A method for photogrammetric measurement includes capturing a sequence of images of an infant lying in a crib and processing the images so as to extract a skeleton of the infant from one or more of the images. A height of the infant is estimated based on the extracted skeleton.

Abstract: A video processing apparatus includes a video analyzer that analyzes video data captured by a surveillance camera, detects an event belonging to a specific category, and outputs a detection result, a display controller that displays, together with a video of the video data, a category setting screen for setting a category of an event included in the video, and a learning data accumulator that accumulates, as learning data together with the video data, category information set in accordance with an operation by an operator to the category setting screen. The video analyzer performs learning processing by using the learning data accumulated in the learning data accumulator.

Abstract: The present subject matter provides technical solutions facing technical problems associated with preserving spatial dimension of images used in CNNs. Transposed convolutional layers may be used to provide improved spatial dimension preservation or reconstruction. In contrast with image interpolation, transposed convolutional layers may use a set of weights to reconstruct input images. When using CNNs for ADAS and AV applications, the transposed convolutional layers may be trained jointly with convolutional layers during the CNN training process. This may provide the ability to use a lower-dimensional representation of input images, while preserving the spatial dimension of images for use in ADAS and AV systems.

Abstract: An inspection system for a welded portion in welding includes a display unit, an acquisition unit that acquires a reference position of a welded portion stored in the storage device, a display control unit that displays, on the display unit, a reference position image that is an image showing a reference position virtually superimposed on an image of an object to be welded in a real space captured by the imaging device and includes a welded portion, and a distance calculation unit that calculates a distance in the real space between the welded portion in the real space displayed on the display unit and the reference position of the welded portion.

Abstract: Aspects of the present invention relate to a control system (1) for identifying at least a first activity performed by a person of interest (POI-n). The control system (1) includes a controller having a processor (20) and a system memory (26). The controller (13) is configured to receive image data (DIMG1, DIMG2) from at least one imaging device (C1, C2). The image data (DIMG1, DIMG2) includes at least one image frame representing an image scene including at least a portion of a person of interest (POI-n). The image data (DIMG1, DIMG2) is analysed to identify a skeletal model (15) of the at least a portion of the person of interest (POI-n); and at least one object of interest (OOI-n) with which the person of interest (POI-n) is interacting. The first activity being performed by the person of interest (POI-n) is identified in dependence on the skeletal model (15) and the identified object of interest (OOI-n).

Abstract: Methods for detecting areas of localized tilt on a sample using imaging reflectometry measurements include obtaining a first image without blocking any light reflected from the sample and obtaining a second image while blocking some light reflected from the sample at the aperture plane. The areas of localized tilt are detected by comparing first reflectance intensity values of pixels in the first image with second reflectance intensity values of corresponding pixels in the second image.

Abstract: A method for selecting a crop score threshold for enhancing tracking of objects in a scene captured in a video sequence is disclosed. A respective track is obtained for two different objects, each track comprising crops of object instances of the objects in in a video sequence, each crop having a crop score and a feature vector. Each track is split into respective more tracklets thereby forming four or more tracklets. For each candidate crop score threshold a respective difference between each tracklet and each other tracklet is determined based on differences between feature vectors of crops having a crop score above the candidate crop score threshold of each tracklet, and each other tracklet. A crop score threshold is selected from the set of crop score thresholds resulting in a maximum difference between the differences between tracklets of different tracks and the differences between tracklets of the same track.

Abstract: Disclosed herein are systems and method for classifying objects in an image using a color-based neural network. A method may include: training a neural network to classify an object in a given image into a color class from a set of color classes; determining, from the set of color classes, a subset of color classes that are anticipated to be in a received input image based on image metadata; generating a matched mask input indicating the subset; inputting both the input image and the matched mask input into the neural network, wherein the neural network is configured to: determine a first semantic embedding of the input image and the matched mask input; outputting a color class associated with a second semantic embedding with a least amount of distance to the first semantic embedding from a plurality of semantic embeddings.

Abstract: Disclosed are a method, an apparatus, a device, and a medium for image recognition via wireless federated learning. The method for image recognition via wireless federated learning includes: obtaining an image to be recognized and an initial image recognition model; adjusting parameters for the initial image recognition model via a preset accelerated mobile federated learning algorithm according to a target momentum factor to obtain a target image recognition model; and recognizing the image to be recognized through the target image recognition model to obtain text information corresponding to the image to be recognized.

Abstract: The present application relates to a method for vehicle hinge point calibration, comprising: acquiring a raw image of a first portion and a second portion hinged to one another of a vehicle; adjusting, stitching and cropping the raw image to obtain respective independent surround-view images of the first portion and the second portion; and at least on the basis of the independent surround-view images of the first portion and the second portion, calculating the angle of rotation and the corresponding translation vector between the first and second portions, and thus calculating the coordinates of the respective hinge points of the first and second portions. Further comprised in the present application are an apparatus for vehicle hinge point calibration, a computer device, and a computer-readable storage medium.

Abstract: Techniques for inferring whether an event is occurring in 3D space based on 2D image data and for maintaining a camera's calibration are disclosed. An image of an environment is accessed. Input is received, where the input includes a 2D rule imposed against a ground plane. The 2D rule includes conditions indicative of an event. A bounding box is generated and encompasses a detected object. A point within the bounding box is projected from a 2D-space image plane of the image into 3D space to generate a 3D-space point. Based on the 3D-space point, a 3D-space ground contact point is generated. That 3D-space ground contact point is reprojected onto the ground plane of the image to generate a synthesized 2D ground contact point. A location of the synthesized 2D ground contact point is determined to satisfy the conditions.