Abstract: A method for 3D human model reconstruction and animation includes receiving a two-dimensional (2D) image of a human, segmenting the 2D image into a foreground with the human and a background without the human in the 2D image, generating a parametric model comprising a pose, a shape, and one or more rigging parameters based on the human in the foreground, predicting a textured three-dimensional (3D) model using implicit surface reconstruction of the human in the foreground, aligning the parametric model and the textured 3D model using a 3D registration, and generating a textured 3D clothed human model based on the aligned parametric model and the predicted textured 3D model.

Abstract: An embodiment includes parsing geographical data into a path graph having a plurality of nodes and edges, and identifying first and second subsets of the nodes as source nodes and destination nodes, respectively. The embodiment generates path data for a candidate delivery route from a source node to a destination node and along an edge between the source and destination nodes. The embodiment processes the path data using first and second evaluation techniques based on respective metrics. The embodiment compares evaluation values from the evaluation techniques to evaluation values associated with another candidate delivery route, and selects the candidate delivery route as a finalized delivery route based on the comparison results. The embodiment then generates a route plan that includes the finalized delivery route.

Abstract: From a set of point data, a set of scattered rays is constructed. From the set of scattered rays, a set of ray slopes is computed. The set of ray slopes is mapped to a corresponding set of trigonometric functions. Using an optimization method, a parameter of the set of trigonometric functions is selected. Using an inverse of the set trigonometric functions, a vehicle mass corresponding to the set of point data is computed. Based on the vehicle mass, a threshold braking distance of a collision avoidance system of the vehicle is adjusted, the threshold braking distance comprising a distance from an object predicted to collide with the vehicle. By braking the vehicle at least the threshold braking distance from the object, a predicted collision between the vehicle and the object is avoided.

Abstract: From a set of point data, a set of scattered rays is constructed. From the set of scattered rays, a set of ray slopes is computed. The set of ray slopes is mapped to a corresponding set of trigonometric functions. Using an optimization method, a parameter of the set of trigonometric functions is selected. Using an inverse of the set trigonometric functions, a vehicle mass corresponding to the set of point data is computed. Based on the vehicle mass, a threshold braking distance of a collision avoidance system of the vehicle is adjusted, the threshold braking distance comprising a distance from an object predicted to collide with the vehicle. By braking the vehicle at least the threshold braking distance from the object, a predicted collision between the vehicle and the object is avoided.

Abstract: Mechanisms are provided for generating an augmented reality representation of a real-world environment. An augmented reality (AR) system receives a captured digital image of the real-world environment and generates an initial estimate of a candidate point specifying an estimated location of an annotation point of a virtual object model within the captured digital image of the real-world environment. An accuracy of the initial estimate is calculated based on a function of characteristics of the annotation point and a function of characteristics of the candidate point and, in response to the evaluation of accuracy indicating that the initial estimate is not accurate, an annotation point location refinement operation is performed to generate a refined candidate point for aligning the annotation point with the captured digital image of the real-world environment. An AR representation of the real-world environment is generated based on the refined candidate point.

Abstract: Embodiments of the disclosure provide methods and systems for positioning a vehicle. The system may include a communication interface configured to receive a set of point cloud data with respect to a scene captured under a first lighting condition by at least one sensor. The system may further include a storage configured to store the set of point cloud data, and a processor. The processor may be configured to simulate illumination of the scene under a second lighting condition different from the first lighting condition based on the set of point cloud data, modify the set of point cloud data based on the simulated illumination of the scene under the second lighting condition, and position the vehicle under the second lighting condition based on the modified set of point cloud data.

Abstract: A method, computer program product, and computer system for finding outliers in multidimensional time series samples. Each time series sample is divided into at least 2 sub samples having equal time duration. At least one prediction model is pre-trained using the sub samples and a prediction result for each sub sample for each prediction model is obtained by executing the pre-trained prediction models with the time series samples as input. A Shapely value corresponding to each prediction result is sub samples for each prediction model to generate multiple clusters of Shapely values for each prediction model. Highest ranking Shapely value outliers are determined from analysis of the multiple clusters. Highest ranking outlier sub samples corresponding to the highest ranking Shapely value outliers are identified.

Abstract: Present disclosure discloses an image processing system and method for manipulating two-dimensional (2D) images of three-dimensional (3D) objects of a predetermined class (e.g., human faces). A 2D input image of a 3D object of the predetermined class is manipulated by manipulating physical properties of the 3D object, such as a 3D shape of the 3D input object, an albedo of the 3D input object, a pose of the 3D input object, and lighting illuminating the 3D input object. The physical properties are extracted from the 2D input image using a neural network that is trained to reconstruct the 2D input image. The 2D input image is reconstructed by disentangling the physical properties from pixels of the 2D input image using multiple subnetworks. The disentangled physical properties produced by the multiple subnetworks are combined into a 2D output image using a differentiable renderer.

Abstract: From a set of point data, a set of scattered rays is constructed. From the set of scattered rays, a set of ray slopes is computed. The set of ray slopes is mapped to a corresponding set of trigonometric functions. Using an optimization method, a parameter of the set of trigonometric functions is selected. Using an inverse of the set trigonometric functions, a vehicle mass corresponding to the set of point data is computed. Based on the vehicle mass, a threshold braking distance of a collision avoidance system of the vehicle is adjusted, the threshold braking distance comprising a distance from an object predicted to collide with the vehicle. By braking the vehicle at least the threshold braking distance from the object, a predicted collision between the vehicle and the object is avoided.

Abstract: The disclosure provides a novel electrocatalytic membrane reactor and use thereof in preparation of high-purity hydrogen. The electrocatalytic membrane reactor adopts an H-shaped electrolytic tank in which a cathode chamber is isolated from an anode chamber through a diaphragm, a membrane electrode is used as an anode, an auxiliary electrode is used as a cathode, a direct-current regulated power supply supplies a constant current, and the flow of a reaction solution is realized through a pump. In the disclosure, electrocatalysis is coupled with a membrane separation function, an oxygen evolution reaction is replaced with an organic electrochemical oxidation reaction in the anode chamber so as to reduce the overpotential of the oxygen evolution reaction, and a hydrogen evolving reaction is performed in the cathode chamber to prepare high-purity hydrogen.

Abstract: A hybrid heat sink includes a solid heat dissipation module and a TS heat dissipation module. The solid heat dissipation module includes a solid substrate and solid fins. The solid substrate has a first side surface and a second side surface opposite to each other. The first side surface is for contacting a heat source, and the solid fins are connected to the second side surface. The TS heat dissipation module includes a TS substrate and TS fins fixed to the TS substrate. A receiving cavity for receiving a phase-change working medium is formed in the TS substrate, condensation reflux cavities are formed in the TS fins and are communicated with the receiving cavity. The TS substrate is fixed at a mounting opening of the solid substrate with a side surface of the TS substrate being exposed to the first side surface for contacting the heat source.

Abstract: A system for flexible regularization and adaptable scaling of an artificial neural network is provided.

Abstract: Methods, computer program products, and/or systems are provided that perform the following operations: obtaining pre-check data associated with specified data nodes; calculating outliers for each specified data node, wherein the outliers are calculated based on a unit of the pre-check data associated with each specified data node; backtracking the calculated outliers for each specified data node through an associated generating data link; selecting one or more data nodes associated with a set of largest outliers; selecting one or more data links associated with the set of largest outliers; and generating potential anomaly indications based on the one or more data nodes selected and the one or more data links selected.

Abstract: The present invention provides an interval error observer-based aircraft engine active fault tolerant control method, and belongs to the technical field of aircraft control. The method comprises: tracking the state and the output of a reference model of an aircraft engine through an error feedback controller; compensating a control system of the aircraft engine having a disturbance signal and actuator and sensor faults through a virtual sensor and a virtual actuator; observing an error between a system with fault of the aircraft engine and the reference model through an interval error observer, and feeding back the error to the error feedback controller; and finally, using a difference between the output of the reference model of the system with fault and the output of the virtual actuator as a control signal to realize active fault tolerant control of the aircraft engine.

Abstract: Present disclosure discloses an image processing system and method for manipulating two-dimensional (2D) images of three-dimensional (3D) objects of a predetermined class (e.g., human faces). A 2D input image of a 3D object of the predetermined class is manipulated by manipulating physical properties of the 3D object, such as a 3D shape of the 3D input object, an albedo of the 3D input object, a pose of the 3D input object, and lighting illuminating the 3D input object. The physical properties are extracted from the 2D input image using a neural network that is trained to reconstruct the 2D input image. The 2D input image is reconstructed by disentangling the physical properties from pixels of the 2D input image using multiple subnetworks. The disentangled physical properties produced by the multiple subnetworks are combined into a 2D output image using a differentiable renderer.

Abstract: Embodiments of the present disclosure disclose a method and a system for training a neural network for generating universal adversarial perturbations. The method includes collecting a plurality of data samples. Each of the plurality of data samples is identified by a label from a finite set of labels. The method includes training a probabilistic neural network for transforming the plurality of data samples into a corresponding plurality of perturbed data samples having a bounded probability of deviation from the plurality of data samples by maximizing a conditional entropy of the finite set of labels of the plurality of data samples conditioned on the plurality of perturbed data samples. The conditional entropy is unknown. The probabilistic neural network is trained based on an iterative estimation of a gradient of the unknown conditional entropy of labels. The method further includes generating the universal adversarial perturbations based on the trained probabilistic neural network.

Abstract: A method, computer program product, and computer system for repairing a Dockerfile. Library versions containing initial version numbers of libraries are extracted from the Dockerfile. A Monte Carlo tree search (MCTS) is executed, using the extracted library versions as input, which generates a tree that includes multiple levels populated with noses. Each node in a level represents the generic library name of a library version in the Dockrerfile and an associated randomly selected version number. At least one of the randomly selected version numbers associated with at least one node in a level differs from the initial version number associated with a versionf. A best successful installation path is selected from the at least one successful installation path. The Dockerfile is repaired by inserting randomly selected version numbers into Dockerfile as replacements for some of the initial version numbers.

Abstract: One or more computer processors encode a plurality of time sequenced global position system (GPS) datapoints onto a grided dimensional area; determine a general trajectory between each time sequenced GPS datapoint in the plurality of encoded time sequenced GPS datapoints and a subsequent encoded time sequenced GPS datapoint; cluster the encoded time sequenced GPS datapoints based on a respective determined trajectory with a plurality of encoded historical GPS datapoints; calculate an azimuth for each encoded time sequenced GPS datapoint in the plurality of time sequenced GPS datapoints utilizing a plurality of adjacent historical GPS datapoints contained within a respective cluster; generate a plurality of interpolated GPS datapoints utilizing calculated azimuths, determined general trajectories, and historical GPS datapoints; and aggregate the generated interpolated GPS datapoints with the plurality of time sequenced GPS datapoints into an interpolated route, wherein each GPS datapoint in the interpolated ro

Abstract: A computer-implemented pre-processed time-delay autoencoder based anomaly detection method are provided for detecting anomalous states of machines arranged in a factory automation (FA) system or a manufacturing production line.