Qualcomm Incorporated

Alcatel Lucent

Alcatel-lucent USA Inc.

Urs Niesen

Jubin Jose

Xinzhou Wu

Venkatesan Nallampatti Ekambaram

Lionel Jacques Garin

Jayakrishnan Unnikrishnan

Junyi Li

Libin Jiang

Lei Zhang

Shrinivas Kudekar

Kapil Gulati

Sudhir Kumar Baghel

Piyush Gupta

Anantharaman Balasubramanian

Chang-Sik Choi

Mohammadali Maddah-Ali

Meghana Bande

Murali Chari

Thomas L Marzetta

Gordon Thomas Wilfong

Arvind Ramanandan

Adiseshu Hari

Avdhut Joshi

Michael John Hamilton

Muryong Kim

Yoga Y Nadaraajan

Amir Salimi

Gautam Sachdeva

Shivam Agarwal

Sree Sesha Aravind Vadrevu

Mohammadali Maddahali

Ramtin Pedarsani

Yu-Chih Huang

Systems and methods for resolving elevation ambiguity include acquiring, using a 1-D horizontal radar antenna array, a radar frame with range and azimuth information, and predicting a target elevation based on the frame by computing a depth map with a plurality of target depths assigned to corresponding azimuth-elevation pairs. Computing the depth map includes processing the radar frame with an encoder-decoder structured deep convolutional neural network (CNN). The CNN may be trained with a dataset including training radar frames acquired in a number of environments, and compensated ground truth depth maps associated with those environments. The compensated ground truth depth maps may be generated by subtracting a ground-depth from a corresponding ground truth depth map. The ground truth depth maps may be acquired with a 2-D range sensor, such as a LiDAR sensor, a 2-D radar sensor, and/or an IR sensor. The radar frame may also include Doppler data.

Resolution of elevation ambiguity in one-dimensional radar processing

Disclosed are techniques for radar-aided single-image three-dimensional (3D) depth reconstruction. In an aspect, at least one processor of an on-board computer of an ego vehicle receives, from a radar sensor of the ego vehicle, at least one radar image of an environment of the ego vehicle, receives, from a camera sensor of the ego vehicle, at least one camera image of the environment of the ego vehicle, receives, from a light detection and ranging (LiDAR) sensor of the ego vehicle, at least one LiDAR image of the environment of the ego vehicle, and generates a depth image of the environment of the ego vehicle based on the at least one radar image, the at least one LiDAR image, and the at least one camera image.

Radar-aided single image three-dimensional depth reconstruction

Disclosed are techniques for performing object detection and tracking. In some implementations, a process for performing object detection and tracking is provided. The process can include steps for obtaining, at a tracking object, an image comprising a target object, obtaining, at the tracking object, a first set of messages associated with the target object, determining a bounding box for the target object in the image based on the first set of messages associated with the target object, and extracting a sub-image from the image. In some approaches, the process can further include steps for detecting, using an object detection model, a location of the target object within the sub-image. Systems and machine-readable media are also provided.

Objection detection using images and message information

Methods, systems, computer-readable media, and apparatuses for radar or LIDAR measurement are presented. Some configurations include transmitting, via a transceiver, a first beam having a first frequency characteristic; calculating a distance between the transceiver and a moving object based on information from at least one reflection of the first beam; transmitting, via the transceiver, a second beam having a second frequency characteristic that is different than the first frequency characteristic, wherein the second beam is directed such that an axis of the second beam intersects a ground plane; and calculating an ego-velocity of the transceiver based on information from at least one reflection of the second beam. Applications relating to road vehicular (e.g., automobile) use are described.

Ego-velocity estimation using radar or LIDAR beam steering

Techniques provided herein are directed toward virtually extending an updated set of output positions of a mobile device determined by a VIO by combining a current set of VIO output positions with one or more previous sets of VIO output positions in such a way that ensure all outputs positions among the various combined sets of output positions are consistent. The combined sets can be used for accurate position determination of the mobile device. Moreover, the position determination further may be based on GNSS measurements.

Position-window extension for GNSS and visual-inertial-odometry (VIO) fusion

A user equipment (UE) in a vehicle (V-UE) broadcasts multi-phased ranging signals with which other entities may determine the range to the V-UE. The multi-phased ranging signals may include a first message, which may be broadcast in the Intelligent Transport System (ITS) spectrum, includes ranging information, such as a source identifier, location information for the broadcasting V-UE, and an expected time of broadcast of the ranging signal. The ranging signal may then be broadcast at the expected time and may include the source identifier. A second message, which be broadcast in the ITS spectrum, may include clock error information for the V-UE. A receiving entity may determine the range to the V-UE based on the time of arrival of the ranging signal and the expected time of transmission, as well as the clock error information. The receiving entity may further generate a position estimate based on the received location information.

Sidelink positioning based on physical ranging signals

In an embodiment, a user equipment (UE) receives, from a fixed reference node, at least one round-trip propagation time (RTT) ranging scheduling message indicating a set of downlink (DL) ranging resource assignments and a set of uplink (UL) ranging resource grants, receives one or more DL ranging signals from the fixed reference node on a first set of resources identified by the set of DL ranging resource assignments, and transmits one or more UL ranging signals to the fixed reference node on a second set of resources identified by the set of UL ranging resource grants.

Ranging between a user equipment and a fixed reference node

A hybrid approach for using reference frames is presented in which a series of anchor frames is used, effectively resetting a global frame upon a trigger event. With each new anchor frame, parameter values for lane boundary estimates (known as lane boundary states) can be recalculated with respect to the new anchor frame. Triggering events may a based on a length of time, distance traveled, and/or an uncertainty value.

Lane mapping and localization using periodically-updated anchor frames

Association algorithms of newly-detected lane boundaries to lane boundaries can be made more robust through the use of generated or 'dummy' states. Different types of dummy states can be used to identify outlier/erroneous detections and/or new, legitimate lane boundaries. Therefore, depending on a type of dummy state a newly-detected lane boundary is associated with, the newly-detected lane boundary can be ignored, or the associated dummy state can be added to the lane boundary states of the filter.

Robust lane-boundary association for road map generation

Techniques are described herein for allowing one or more vehicles or radar systems in an environment to passively detect radar signals from other vehicles or other radar systems and determine spatial parameters of objects based on the passively received radar signals. A primary vehicle (or user equipment (UE) associated with the primary vehicle) may be configured to receive one or more radar signals from one or more secondary vehicles (or UEs associated with the secondary vehicles). The primary vehicle may be configured to determine one or more spatial parameters of the secondary vehicle based on the passively received radar signals. In some cases, the primary vehicle may receive an indication that identifies at least some communication resources to be used by the secondary vehicle to transmit the radar signals. The primary vehicle may determine one or more driving operations based on determining the spatial parameter.

System and method to use reflected doppler radar signals to locate a second mobile device

Techniques and systems are provided for determining one or more sizes of one or more objects. For example, a bounding region identifying a first object detected in an image can be obtained. A map including map points can also be obtained. The map points correspond to one or more reference locations in a three-dimensional space. The bounding region identifying the first object can be associated with at least one map point of the map points included in the map. Using the bounding region and the at least one map point, an estimated three-dimensional position and an estimated size of the first object detected in the image can be determined. In some examples, other information can be used to estimate the estimated three-dimensional position and an estimated size of the first object, such as radar information and/or other information.

Object size estimation using camera map and/or radar information

The present disclosure provides methods and systems for radar detection. A radar device may obtain prior information about known targets including an azimuth angle of each known target. The radar device may generate an association aware transmit beam pattern toward at least a subset of the known targets based on the prior information. The radar device may detect targets from reflected beams of the association aware transmit beam pattern. Generating the association aware transmit beam pattern may include generating a partial ambiguity graph for the known targets, the graph including at least one edge connecting two known targets wherein a difficulty of disambiguating the two known targets is greater than a threshold. The graph also includes at least two known targets that are not connected by an edge.

Association aware radar beamforming

Methods, apparatuses, and computer-readable media are described. In one example, a method of controlling a vehicle comprises: receiving, using one or more sensors, a first set of measurements of a set of physical attributes of the vehicle in a motion; determining, based on a motion data model that defines a set of relationships among the set of physical attributes of the vehicle in the motion and based on the first set of measurements, a set of expected measurements of the set of physical attributes; determining whether to use an entirety of the first set of measurements to control an operation of the vehicle based on comparing the first set of measurements and the set of expected measurements; and responsive to determining not to use the entirety of the first set of measurements, controlling the operation of the vehicle based on a second set of measurements.

Growing community of inventors

Summit, NJ, United States of America

Urs Niesen