Qualcomm Incorporated

Michael John Hamilton

Sundar Subramanian

Volodimir Slobodyanyuk

Amin Ansari

Radhika Dilip Gowaikar

Junyi Li

Kapil Gulati

Makesh Pravin John Wilson

Shantanu Chaisson Sanyal

Xinzhou Wu

Ahmed Kamel Sadek

Urs Niesen

Ravi Teja Sukhavasi

Jayakrishnan Unnikrishnan

Bence Major

Daniel Hendricus Franciscus Fontijne

A device for processing image data is disclosed. The device can obtain a radar point cloud and one or more frames of camera data. The device can determine depth estimates of one or more pixels of the one or more frames of camera data. The device can generate a pseudo lidar point cloud using the depth estimates of the one or more pixels of the one or more frames of camera data, wherein the pseudo lidar point cloud comprises a three-dimensional representation of at least one frame of the one or more frames of camera data. The device can determine one or more object bounding boxes based on the radar point cloud and the pseudo lidar point cloud.

Radar and camera data fusion

In some aspects, a system may receive, from a first one-dimensional radar array, first information based at least in part on first reflections associated with an azimuthal plane. The system may further receive, from a second one-dimensional radar array, second information based at least in part on second reflections associated with an elevation plane. Accordingly, the system may detect an object based at least in part on the first information and may determine an elevation associated with the object based at least in part on the second information. Numerous other aspects are described.

Object detection and ranging using one-dimensional radar arrays

Disclosed are techniques for employing deep learning to analyze radar signals. In an aspect, an on-board computer of a host vehicle receives, from a radar sensor of the vehicle, a plurality of radar frames, executes a neural network on a subset of the plurality of radar frames, and detects one or more objects in the subset of the plurality of radar frames based on execution of the neural network on the subset of the plurality of radar frames. Further, techniques for transforming polar coordinates to Cartesian coordinates in a neural network are disclosed. In an aspect, a neural network receives a plurality of radar frames in polar coordinate space, a polar-to-Cartesian transformation layer of the neural network transforms the plurality of radar frames to Cartesian coordinate space, and the neural network outputs the plurality of radar frames in the Cartesian coordinate space.

Radar deep learning

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

Embodiments provided herein allow for identification of one or more regions of interest in a radar return signal that would be suitable for selected application of super-resolution processing. One or more super-resolution processing techniques can be applied to the identified regions of interest. The selective application of super-resolution processing techniques can reduce processing requirements and overall system delay. The output data of the super-resolution processing can be provided to a mobile computer system. The output data of the super-resolution processing can also be used to reconfigure the radar radio frequency front end to beam form the radar signal in region of the detected objects. The mobile computer system can use the output data for implementation of deep learning techniques. The deep learning techniques enable the vehicle to identify and classify detected objects for use in automated driving processes. The super-resolution processing techniques can be implemented in analog and/or digital circuitry.

Super-resolution enhancement techniques for radar

Certain aspects provide a method for radar detection by an apparatus. The method including transmitting a radar waveform in transmission time intervals (TTIs) to perform detection of a target object. The method further includes varying the radar waveform across TTIs based on one or more radar transmission parameters.

Radar transmission time interval randomized radar transmissions

Certain aspects provide a method for radar detection by an apparatus. The method generally includes transmitting a radar waveform in sets of transmission time intervals (TTIs), using a common set of radar transmission parameters in each set of TTIs, to perform detection of a target object, varying at least one of the common set of radar transmission parameters between sets of TTIs, and identifying interfering signals based on observed changes in monitored parameters of received signals across sets of TTIs due to the varying.

Multi-radar coexistence using slow rate interference identification and suppression

Apparatus and methods are disclosed for communicating between distributed automotive sensors, including radar sensors, wherein sensors transmit a synchronization (SYNC) signal, each SYNC signal transmitted via a substantially equal-length fiber optic link corresponding with each sensor. A central node receives the SYNC signals via the fiber optic links corresponding with each of the sensors and determines a master SYNC signal based on the received SYNC signals. The central node then transmits the master SYNC signal via the fiber optic links to the sensors, which receive the master SYNC signal and transmit, via fiber optic link, sensor data synchronized in accordance with the master SYNC signal. The synchronized sensor data are received at the central node and coherently aggregated, and transmitted to a compute node for post-processing. For radar data, the post-processing may include determination of an angular position of an object within detection range of at least two radar sensors.

Distributed automotive radar architecture

Certain aspects of the present disclosure provide techniques for radar detection by an apparatus. In certain aspects a method for radar detection by an apparatus includes selecting one or more radar transmission parameters based on a reference time, wherein the reference time is common to at least a group of vehicles. The method further includes performing radar detection using the selected radar transmission parameters and the reference time.

Growing community of inventors

San Diego, CA, United States of America

Michael John Hamilton