Abstract: A system and method for detecting motion with respect an object includes providing a processor and at least one video sensor; synchronizing the at least one video sensor to a dynamic event associated with the object; recording, by the at least one video sensor, a plurality of data sets including the object, each data set of the plurality of data sets being synchronized with respect to the dynamic event associated with the object; averaging, by the processor, the plurality of data sets to provide an averaged synchronized data set; and calculating, by the processor, a motion with respect to the object based on the averaged synchronized data set.

Abstract: A system includes: a depth module including an encoder and a decoder and configured to: receive a first image from a first time from a camera; and based on the first image, generate a depth map including depths between the camera and objects in the first image; a pose module configured to: generate a first pose of the camera based on the first image; generate a second pose of the camera for a second time based on a second image; and generate a third pose of the camera for a third time based on a third image; and a motion module configured to: determine a first motion of the camera between the second and first times based on the first and second poses; and determine a second motion of the camera between the second and third times based on the second and third poses.

Abstract: This disclosure relates generally to system and method for forecasting location of target in monocular first person view. Conventional systems for location forecasting utilizes complex neural networks and hence are computationally intensive and requires high compute power. The disclosed system includes an efficient and light-weight RNN based network model for predicting motion of targets in first person monocular videos. The network model includes an auto-encoder in the encoding phase and a regularizing layer in the end helps us get better accuracy. The disclosed method relies entirely just on detection bounding boxes for prediction as well as training of the network model and is still capable of transferring zero-shot on a different dataset.

Abstract: The in-vehicle device transfers the selected image selected by the user from the photographed images stored in the storage device and the selected vehicle state, which is the vehicle state when the selected image is captured, to the form terminal. The portable device transmits the selected image and the selected vehicle state to the image management server only when the transmission is permitted by the user. The image management server determines whether or not the selected image is a rare image using the selected image and the selected vehicle state, and stores the selected image in the image storage device when the selected image is determined to be a rare image.

Abstract: A method for photoacoustic imaging is performed by measuring RF time samples from transducer elements, and reconstructing an estimated initial pressure image from the measured RF time samples and a pre-calculated forward model matrix. The reconstructing involves minimizing the difference between the pre-calculated forward model matrix applied to the image estimate and the measured RF time samples by implementing a superiorized modified conjugate gradient least squares algorithm to minimize the total variation norm. The model matrix factors each of the transducer elements into sub-wavelength mathematical elements.

Abstract: A method detects false positives of an image-processing device of a camera located on-board a vehicle and that delivers data for modelling a marking line. The method includes: determining a first distance from the vehicle to the marking line at a first instant in time, determining a second distance from the vehicle to the marking line at a second instant, computing an initial discontinuity, in which an initial discontinuity equal to the absolute value of the difference between the first distance and the second distance is computed, and detecting a false initial-discontinuity positive, in which the initial discontinuity is compared to a predefined initial low threshold.

Abstract: A method and system for providing a mask image including receiving medical image data including a temporal dimension and generating a frequency data set including data points with in each case one frequency value by applying a Fourier transform to the image data. The Fourier transform is applied at least along the temporal dimension. The frequency data set is segmented into two sub-areas based on at least a frequency threshold value. The mask image is generated by applying an inverse Fourier transform to the first and/or the second sub-area of the frequency data set. The generated mask image is provided.

Abstract: Space-based ground-imaging lidar has become increasingly feasible with recent advances in compact fiber lasers and single-photon-sensitive Geiger-mode detector arrays. A challenge with such a system is imperfect pointing knowledge caused by angular jitter, exacerbated by long distances between the satellite and the ground. Unless mitigated, angular jitter blurs the 3D lidar data. Unfortunately, using mechanical isolation, advanced IMUs, star trackers, or auxiliary passive optical sensors to reduce the error in pointing knowledge increases size, weight, power, and/or cost. Here, the 2-axis jitter time series is estimated from the lidar data. Simultaneously, a single-surface model of the ground is estimated as nuisance parameters. Expectation Maximization separates signal and background detections, while maximizing the joint posterior probability density of the jitter and surface states.

Abstract: An image processing apparatus (10) is disclosed that comprises a processor arrangement (16) adapted to receive image data corresponding to a region of interest (1) of a patients cardiovascular system, said image data comprising a temporal sequence (15) of intravascular ultrasound images acquired (150) at different phases of at least one cardiac cycle of said patient, said intravascular ultrasound images imaging overlapping volumes of the patient's cardiovascular system; implement a spatial reordering process of said temporal sequence of intravascular ultrasound images by evaluating the image data to select at least one spatial reference (6, Vref) associated with said temporal sequence of intravascular ultrasound images; estimating a distance to the at least one spatial reference for each of the intravascular ultrasound images of said temporal sequence; and reordering said temporal sequence of intravascular ultrasound images into a spatial sequence of intravascular ultrasound images based on the estimated dist

Abstract: There is provided a method and a system for customized image denoising with interpretability. A deep neural network (NN) is trained to denoise an image on a training dataset including pairs of noisy and corresponding clean images acquired from an imaging apparatus, where during the training a structured covariance score (SCS) indicative of a performance of the deep NN in recovering content of corresponding clean images relative to the denoised image is determined based on sparse conditional correlations. A test noisy image is received and denoised by the deep NN. A user feedback score indicative of user satisfaction of the denoising is obtained. A quality parameter is obtained based on the SCS and a quality metric indicative of denoised image quality is obtained from a pretrained NN, and compared with the user feedback score. If the SCS is above the user feedback score, the deep NN is provided for denoising.