Abstract: Embodiments include a method for object detection in a Light Detection And Ranging (LiDAR) point cloud, the method comprising: placing, by a navigation system, a plurality of anchor points in a two-dimensional Bird's Eye View (BEV) of spatial points represented in a segmented ground surface representation of objects detected by a LiDAR system; extracting, by the navigation system, one or more features from the two-dimensional BEV of the spatial points; proposing, by the navigation system, one or more regions of the two-dimensional BEV of the spatial points for object detection; and performing, by the navigation system, object detections on anchor points of the plurality of anchor points in the proposed one or more regions of the two-dimensional BEV of the spatial points.

Abstract: Embodiments include a method for ground surface segmentation on sparse Light Detection And Ranging (LiDAR) point clouds comprising: reading a LiDAR point cloud from a LiDAR sensor, the LiDAR point cloud comprising data representing one or more objects in physical surroundings detected by the LiDAR sensor; voxelizing the LiDAR point cloud to produce a three-dimensional representation of each of the one or more objects; constructing a maximum height map from the three-dimensional representation of each of the one or more objects, the maximum height map comprising a two-dimensional mapping of spatial points representing each of the one or more objects; performing minimum filtering on the spatial points of the maximum height map; and classifying each spatial point as a ground point or a non-ground point based on the minimum filtering of each spatial point.

Abstract: Systems and methods for generating simulated LiDAR data using RADAR and image data are provided. An algorithm is trained using deep-learning techniques such as loss functions to generate simulated LiDAR data using RADAR and image data. Once trained, the algorithm can be implemented in a system, such as a vehicle, equipped with RADAR and image sensors in order to generate simulated LiDAR data describing the system's environment. The simulated LiDAR data may be used by a vehicle control system to determine, generate, and implement modified driving operations.

Abstract: Embodiments include a method for pruning anchor points from a feature map generated from a Light Detections And Ranging (LiDAR) point cloud, the method comprising: receiving, by a navigation system, a LiDAR point cloud from a LiDAR sensor, the LiDAR point cloud comprising data representing one or more objects in physical surroundings detected by the LiDAR sensor; extracting, by the navigation system, a feature map from the LiDAR point cloud, the feature map comprising a plurality of anchor points, each anchor point defined by an anchor box; smoothing, by the navigation system, the extracted feature map; determining, by the navigation system, density of pixels within the anchor box of each anchor point; and pruning, by the navigation system, anchor points from the feature map based on a plurality of factors related to the determined density of pixels within the box of each anchor point.

Abstract: Systems and methods for generating simulated LiDAR data using RADAR and image data are provided. An algorithm is trained using deep-learning techniques such as loss functions to generate simulated LiDAR data using RADAR and image data. Once trained, the algorithm can be implemented in a system, such as a vehicle, equipped with RADAR and image sensors in order to generate simulated LiDAR data describing the system's environment. The simulated LiDAR data may be used by a vehicle control system to determine, generate, and implement modified driving operations.

Abstract: Embodiments include a method for pruning anchor points from a feature map generated from a Light Detections And Ranging (LiDAR) point cloud, the method comprising: receiving, by a navigation system, a LiDAR point cloud from a LiDAR sensor, the LiDAR point cloud comprising data representing one or more objects in physical surroundings detected by the LiDAR sensor; extracting, by the navigation system, a feature map from the LiDAR point cloud, the feature map comprising a plurality of anchor points, each anchor point defined by an anchor box; smoothing, by the navigation system, the extracted feature map; determining, by the navigation system, density of pixels within the anchor box of each anchor point; and pruning, by the navigation system, anchor points from the feature map based on a plurality of factors related to the determined density of pixels within the box of each anchor point.

Abstract: Embodiments include a method for ground surface segmentation on sparse Light Detection And Ranging (LiDAR) point clouds comprising: reading a LiDAR point cloud from a LiDAR sensor, the LiDAR point cloud comprising data representing one or more objects in physical surroundings detected by the LiDAR sensor; voxelizing the LiDAR point cloud to produce a three-dimensional representation of each of the one or more objects; constructing a maximum height map from the three-dimensional representation of each of the one or more objects, the maximum height map comprising a two-dimensional mapping of spatial points representing each of the one or more objects; performing minimum filtering on the spatial points of the maximum height map; and classifying each spatial point as a ground point or a non-ground point based on the minimum filtering of each spatial point.

Abstract: Embodiments include a method for object detection in a Light Detection And Ranging (LiDAR) point cloud, the method comprising: placing, by a navigation system, a plurality of anchor points in a two-dimensional Bird's Eye View (BEV) of spatial points represented in a segmented ground surface representation of objects detected by a LiDAR system; extracting, by the navigation system, one or more features from the two-dimensional BEV of the spatial points; proposing, by the navigation system, one or more regions of the two-dimensional BEV of the spatial points for object detection; and performing, by the navigation system, object detections on anchor points of the plurality of anchor points in the proposed one or more regions of the two-dimensional BEV of the spatial points.

Abstract: Step detection accuracy in mobile devices is increased by determining whether swinging is taking place. According to the invention, swinging can be detected using threshold detection, Eigen analysis, hybrid frequency analysis, and/or gyroscope-based analysis, for example. The determination that swinging is (or may be) occurring can impact how the mobile device reports detected steps for step detection. A count of missteps and/or a level of certainty, based on swing detection, can be provided with a step count.

Abstract: Disclosed is an apparatus and method for power efficient processor scheduling of features. In one embodiment, features may be scheduled for sequential computing, and each scheduled feature may receive a sensor data sample as input. In one embodiment, scheduling may be based at least in part on each respective feature's estimated power usage. In one embodiment, a first feature in the sequential schedule of features may be computed and before computing a second feature in the sequential schedule of features, a termination condition may be evaluated.

Abstract: An apparatus and method for providing an improved heading estimate of a mobile device in a vehicle is presented. First, the mobile device determines if it is mounted in a cradle attached to the vehicle; if so, inertia sensor data may be valid. While in a mounted stated, the mobile device determines whether it has been rotated in the cradle; if so, inertia sensor data may no longer be reliable and a recalibration to determine a new relative orientation between the vehicle and the mobile device is needed. If the mobile device is mounted and not recently rotated, heading data from multiple sensors (e.g., GPS, gyroscope, accelerometer) may be computed and combined to form the improved heading estimate. This improved heading estimate may be used to form an improved velocity estimate. The improved heading estimate may also be used to compute a bias to correct a gyroscope.

Abstract: Aspects of the disclosure relate to computing technologies. In particular, aspects of the disclosure relate to mobile computing device technologies, such as systems, methods, apparatuses, and computer-readable media for improving orientation data. In some embodiments, a magnetic vector filter receives magnetometer data from a magnetometer and gyroscope data from a gyroscope and determines the magnetic vector using the magnetometer data and the gyroscope data in the magnetic vector filter. In other embodiments, a gravity vector filter receives accelerometer data and gyroscope data and determines the gravity vector using the accelerometer data and the gyroscope data in the gravity vector filter.

Abstract: System and methods are disclosed to use information available on the state of mobile devices in a heuristics-based approach to improve motion state detection. In one or more embodiments, information on the WiFi connectivity of mobile devices may be used to improve the detection of the in-transit state. The WiFi connectivity information may be used with sensor signal such as accelerometer signals in a motion classifier to reduce the false positives of the in-transit state. In one or more embodiments, information that a mobile device is connected to a WiFi access point (AP) may be used as heuristics to reduce the probability of falsely classifying the mobile device in the in-transit state when mobile device is actually in the hand of a relatively stationary user. Information on the battery charging state or the wireless connectivity of the mobile devices may also be used to improve the detection of in-transit state.

Abstract: Aspects of the disclosure relate to computing technologies. In particular, aspects of the disclosure relate to mobile computing device technologies, such as systems, methods, apparatuses, and computer-readable media for scheduling an execution of a task, such as a non-real time, non-latency sensitive background task on a computing device. In one implementation, the technique includes detecting a first state of a device, wherein the first state of the device is associated with a first power level and a first task, wherein the first power level is at least partially based on power consumption of a first task, determining that the first power level associated with the first state is above a threshold, and in response to determining that the first power level associated with the first state is above the threshold, and scheduling an execution of a second task on the device, wherein the second task is associated with automatically collecting of calibration data using at least one sensor.

Abstract: Aspects of the disclosure relate to computing technologies. In particular, aspects of the disclosure relate to mobile computing device technologies, such as systems, methods, apparatuses, and computer-readable media for improving calibration data by increasing the diversity of orientations used for generating the calibration data. In one embodiment, the computing device receives a plurality of calibration measurements associated with one or more sensors of a device, determines a degree to which the plurality of calibration measurements were captured at different orientations of the device, and determines, based on the degree, whether to update one or more calibration parameters.

Abstract: Aspects of the disclosure relate to computing technologies. In particular, aspects of the disclosure relate to mobile computing device technologies, such as systems, methods, apparatuses, and computer-readable media to improve the calibration data by taking into account the effects of change in temperature on motion sensors. For instance, different levels of error may be associated with a motion sensor at different temperature levels. In one implementation, the sensor measurement data associated with the various orientations at a temperature is used in determining the calibration data for that temperature.

Abstract: A user equipment (UE) controls power consumption of the UE based on mobility information and channel conditions experienced by the UE. In one instance, the UE determines its level of mobility based on a Doppler frequency spread of received communications. The UE disables a motion sensor when the level of mobility is above a first threshold. The UE then controls the communications based on the motion sensor and channel conditions experienced by the UE when the level of mobility is below the first threshold.

Abstract: System and methods are disclosed to use information available on the state of mobile devices in a heuristics-based approach to improve motion state detection. In one or more embodiments, information on the WiFi connectivity of mobile devices may be used to improve the detection of the in-transit state. The WiFi connectivity information may be used with sensor signal such as accelerometer signals in a motion classifier to reduce the false positives of the in-transit state. In one or more embodiments, information that a mobile device is connected to a WiFi access point (AP) may be used as heuristics to reduce the probability of falsely classifying the mobile device in the in-transit state when mobile device is actually in the hand of a relatively stationary user. Information on the battery charging state or the wireless connectivity of the mobile devices may also be used to improve the detection of in-transit state.

Abstract: Step detection accuracy in mobile devices is increased by determining whether swinging is taking place. According to the invention, swinging can be detected using threshold detection, Eigen analysis, hybrid frequency analysis, and/or gyroscope-based analysis, for example. The determination that swinging is (or may be) occurring can impact how the mobile device reports detected steps for step detection. A count of missteps and/or a level of certainty, based on swing detection, can be provided with a step count.

Abstract: Methods, systems, computer-readable media, and apparatuses for tap detection in a mobile device are presented. In some embodiments, the method may comprise storing, by a mobile device, a first data sample from an accelerometer sensor and a second data sample from a gyroscope sensor. Additionally, the method may comprise processing a plurality of data samples. The plurality of data samples can include the first data sample or the second data sample. Optionally, in one embodiment, the method may comprise suppressing a tap that has been classified as a false detection based on at least one of the plurality of data samples. Subsequently, the method may comprise determining an occurrence of a tap at a mobile device based on the results of the processing.