Abstract: The disclosure relates to technology for detecting and tracking eye gaze. An apparatus comprises a visible wavelength camera, an infrared (IR) camera, and one or more processors. The one or more processors are configured to generate a three-dimensional (3D) point cloud of a person's face from IR data captured from the IR camera, generate a two-dimensional image of the person's face from visible wavelength data captured from the visible wavelength camera, and detect a symmetry plane of the person's face based on the 3D point cloud and the two-dimensional image. The symmetry plane divides the 3D point cloud into two portions. The one or more processors are further configured to reconstruct the 3D point cloud based on the symmetry plane, and track eye gaze of the person's face based on the reconstructed 3D point cloud.

Abstract: The present technology discloses a federated learning network including a server and multiple client devices. The server receives a set of parameters of a local machine-learning model from each client device in a subset of the multiple client devices. The set of parameters are combined from each of the client devices in the subset to generate an integrated set of parameters. The server then calculates a parameter difference between the integrated set of parameters and the set of parameters for each client device in the subset. Feedback is sent by the server to each client device in the subset. The feedback is applied during backpropagation of the client. If the local parameters of a client are determined to be invalid for a number of times, the client will be set as an outlier.

Abstract: The disclosure relates to technology for acceleration of GPUs in cloud. Instructions for a computational task are accessed. An allocation of data and instructions is calculated based on the data, the instructions, and dynamic GPU resources. The data and the instructions are provided to the GPUs in accordance with the allocation, which includes scheduling a set of instructions for parallel computation of an operation of the computational task on multiple sub-matrices of a data matrix. Separate portions of information are stored into corresponding different regions of non-transitory memory of a processor core to provide concurrent access to the multiple sub-matrices to the processor core. Each sub-matrix corresponds to a portion of the data matrix for which an operation of the computational task is to be performed. Each sub-matrix contains an element in the data matrix in common with another sub-matrix of the data matrix.

Abstract: A method implemented by an agent monitoring device, comprises obtaining, by a sensor of the agent monitoring device, sensor data over a period of time, the sensor data describing a characteristic of an agent associated with the agent monitoring device, determining output data for the sensor based on the sensor data using a learning model, determining a sensor condition for the sensor, determining that a power level of a battery of the agent monitoring device meets the pre-defined power level, determining whether the output data meets the threshold value of the sensor condition in response to the power level of the battery having reached the pre-defined power level, and uploading an indication of the output data to at least one of a cloud server or a representative device in response to the output data for the sensor having met the threshold value of the sensor condition.

Abstract: The disclosure relates to technology for detecting and tracking eye gaze. An apparatus comprises a visible wavelength camera, an infrared (IR) camera, and one or more processors. The one or more processors are configured to generate a three-dimensional (3D) point cloud of a person's face from IR data captured from the IR camera, generate a two-dimensional image of the person's face from visible wavelength data captured from the visible wavelength camera, and detect a symmetry plane of the person's face based on the 3D point cloud and the two-dimensional image. The symmetry plane divides the 3D point cloud into two portions. The one or more processors are further configured to reconstruct the 3D point cloud based on the symmetry plane, and track eye gaze of the person's face based on the reconstructed 3D point cloud.

Abstract: A computer implemented method for self-learning of a control system. The method includes creating an initial knowledge base. The method learns first principles using the knowledge base. The method creates initial control commands derived from the knowledge base. The method generates constraints for the control commands. The method performs constrained reinforcement learning by executing the control commands with the constraints and observing feedback to improve the control commands. The method enriches the knowledge base based on the feedback.

Abstract: A computer-implemented method of answering questions comprises: receiving, by one or more processors, a query; based on the query, generating, by the one or more processors, a matrix; based on the matrix, modifying, by the one or more processors, a dynamic memory; based on the matrix, determining, by the one or more processors, a first response from the dynamic memory; based on the matrix, determining, by the one or more processors, a second response from a database; based on the first response and the second response, determining, by the one or more processors, a third response; and in response to the query, providing, by the one or more processors, the third response.

Abstract: A computer implemented method for self-learning of a control system. The method includes creating an initial knowledge base. The method learns first principles using the knowledge base. The method creates initial control commands derived from the knowledge base. The method generates constraints for the control commands. The method performs constrained reinforcement learning by executing the control commands with the constraints and observing feedback to improve the control commands. The method enriches the knowledge base based on the feedback.

Abstract: Methods, apparatus, and systems are provided for integrated driver expression recognition and vehicle interior environment classification to detect driver condition for safety. A method includes obtaining an image of a driver of a vehicle and an image of an interior environment of the vehicle. Using a machine learning method, the images are processed to classify a condition of the driver and of the interior environment of the vehicle. The machine learning method includes general convolutional neural network (CNN) and CNN with adaptive filters. The adaptive filters are determined based on influence of filters. The classification results are combined and compared with predetermined thresholds to determine if a decision can be made based on existing information. Additional information is requested by self-motivated learning if a decision cannot be made, and safety is determined based on the combined classification results. A warning is provided to the driver based on the safety determination.

Abstract: Various embodiments include systems and methods structured to provide recognition of an object in an image using a learning module trained using decomposition of the object into components in a number of training images. The training can be based on an overall objectness score of the object, an objectness score of each component of the object, a pose of the object, and a pose of each component of the object for each training image input. Additional systems and methods can be implemented in a variety of applications.

Abstract: Methods, apparatus, and systems are provided for integrated driver expression recognition and vehicle interior environment classification to detect driver condition for safety. A method includes obtaining an image of a driver of a vehicle and an image of an interior environment of the vehicle. Using a machine learning method, the images are processed to classify a condition of the driver and of the interior environment of the vehicle. The machine learning method includes general convolutional neural network (CNN) and CNN with adaptive filters. The adaptive filters are determined based on influence of filters. The classification results are combined and compared with predetermined thresholds to determine if a decision can be made based on existing information. Additional information is requested by self-motivated learning if a decision cannot be made, and safety is determined based on the combined classification results. A warning is provided to the driver based on the safety determination.

Abstract: Various embodiments include systems and methods structured to provide recognition of an object in an image using a learning module trained using decomposition of the object into components in a number of training images. The training can be based on an overall objectness score of the object, an objectness score of each component of the object, a pose of the object, and a pose of each component of the object for each training image input. Additional systems and methods can be implemented in a variety of applications.

Abstract: A computer-implemented method of answering questions comprises: receiving, by one or more processors, a query; based on the query, generating, by the one or more processors, a matrix; based on the matrix, modifying, by the one or more processors, a dynamic memory; based on the matrix, determining, by the one or more processors, a first response from the dynamic memory; based on the matrix, determining, by the one or more processors, a second response from a database; based on the first response and the second response, determining, by the one or more processors, a third response; and in response to the query, providing, by the one or more processors, the third response.

Abstract: A computer-implemented method of controlling an autonomous system comprises: accessing, by one or more processors, sensor data that includes information regarding an area; disregarding, by the one or more processors, a portion of the sensor data that corresponds to objects outside of a region of interest; identifying, by the one or more processors, a plurality of objects from the sensor data; assigning, by the one or more processors, a priority to each of the plurality of objects; based on the priorities of the objects, selecting, by the one or more processors, a subset of the plurality of objects; generating, by the one or more processors, a representation of the selected objects; providing, by the one or more processors, the representation to a machine learning system as an input; and based on an output from the machine learning system resulting from the input, controlling the autonomous system.

Abstract: A method exploits user labels in image segmentation. The user labels are propagated with respect to image intensity information. Propagated user labels are included in a cost function of level set evolution. The level set represents a probability of the object segment.

Abstract: A method exploits user labels in image segmentation. The user labels are propagated with respect to image intensity information. Propagated user labels are included in a cost function of level set evolution. The level set represents a probability of the object segment.