Abstract: Certain aspects of the present disclosure provide techniques and apparatus for autonomous driving. In an example method, sensor data associated with an external environment of a vehicle is obtained. A set of prompts for input to a machine learning model trained to assist with autonomous driving for the vehicle is selected via a manager, based at least in part on the sensor data. A set of output tokens corresponding to the set of prompts is generated using the trained machine learning model and the set of prompts. The set of output tokens is associated with at least one of (i) one or more adversarial objects in the environment or (ii) one or more safety events in the environment. Information associated with the set of output tokens is output via the manager for a user of the vehicle.

Abstract: A user equipment (UE) for vehicle-to-everything (V2X) transmission is provided. The UE includes: at least one transceiver; at least one memory; and at least one processor, coupled with the at least one transceiver and the at least one memory, the at least one processor configured to: detect a second position of a second UE with respect to a first position of the UE; generate an alert time window based on the first position of the UE and the second position of the second UE; and schedule at least one transmission from the UE to the second UE based on the alert time window.

Abstract: A computing system of a device may include processing circuitry that determines, based on the sensor data generated by one or more sensors of the device that captures an environment around the device, attributes of external objects in the environment. The processing circuitry may determine priority levels of the external objects based on the attributes of the external objects, wherein the priority levels of the external objects are indicative of a likelihood of the external objects interfering with the device. The processing circuitry may segment the sensor data into a plurality of portions of the sensor data associated with regions of the environment based on the priority levels of the external objects in the regions of the environment. The processing circuitry may adaptively format the plurality of portions of the sensor data based on the priority levels of the external objects in the regions of the environment.

Abstract: Systems and techniques are described herein for attack detection. For instance, a process can include determining, from a received query, a first key and value pair for an application programming interface (API); analyzing the first key and value pair to determine whether the first key and value pair are associated with an attack; and based on a determination that the first key and value pair are not associated with an attack, calling the API using the first key and value pair.

Abstract: Techniques are discussed herein for transmission of location information by a user equipment (UE) to other UEs. A UE receives Satellite Positioning System (SPS) signals and determines whether the SPS signals are reliable. The UE determines a location estimate to be transmitted to other UEs using the SPS signals if the SPS signals are determined to be reliable and using non-SPS information if the SPS signals are determined to be not reliable. The location information is transmitted to other UEs in a message that includes an indication of the source of information used to generate the location estimate. A UE that receives the message may determine its location estimate based, at least in part, on the indication of the source of information, e.g., by determining whether SPS signals are reliable based, at least in part, on the indication of the source of information received in the message.

Abstract: Disclosed are techniques for wireless communication. In an aspect, a first wireless device may receive, from a second wireless device, a first message, the first message comprising an advertised location of the second wireless device (910). The wireless device may determine an estimated location of the second wireless device (920). The wireless device may detect a discrepancy between the advertised location of the second wireless device and the estimated location of the second wireless device (930). The wireless device may send, to at least one other wireless device, a second message, the second message indicating that there is a discrepancy between the advertised location of the second wireless device and the estimated location of the second wireless device (940).

Abstract: Systems and techniques are described herein for attack mitigation. For instance, a process can include allocating a first masked gadget, of a plurality of masked gadgets, for executing a first machine learning (ML) model; allocating a second masked gadget, of the plurality of masked gadgets, for executing a second ML model; and using the first masked gadget with the first ML model concurrently with the second masked gadget with the second ML model.

Abstract: The present disclosure provide techniques for object detection. A method may include extracting a plurality of vehicle-to-everything (V2X) features from a first plurality of V2X messages, wherein the first plurality of V2X messages comprise information about one or more first objects in a scene during a first time period; generating a bird's eye view (BEV) representation of the scene based on, at least, projecting the plurality of V2X features from a V2X space to a BEV space; and detecting, based on the BEV representation, at least a first object among the one or more first objects in the scene.

Abstract: Systems and techniques are described herein for attack detection. For instance, a process can include determining, from a received query, a first key and value pair for an application programming interface (API); analyzing the first key and value pair to determine whether the first key and value pair are associated with an attack; and based on a determination that the first key and value pair are not associated with an attack, calling the API using the first key and value pair.

Abstract: Systems and techniques are provided for securing data. For instance, a process can include: obtaining a session key based on a security handshake operation with a sensor controller; obtaining sensor data from a sensor; scrambling the sensor data based on the session key to generate scrambled sensor data; and outputting the scrambled sensor data to the sensor controller for processing.

Abstract: Systems and techniques are described herein for perturbation detection. For example, a computing device can produce, using a first model based on input data, first output data. The computing device can produce, using a second model based on the input data, second output data. The computing device can determine, based on the first output data and the second output data, a consistency score indicating a consistency between the first output data and the second output data. The computing device can determine, based on the consistency score being less than a consistency score threshold, the input data comprises a perturbation.

Abstract: A processing circuit is coupled to imaging devices through multiple data communication links. Each data communication link couples at least one imaging device with the processing circuit. The processing circuit is configured to determine when a change is required in a first data protection configuration that protects first image data transmitted over a first data communication link, and determine a second data protection configuration to be used when second image data is transmitted from a first imaging device over the first data communication link to an image processing circuit. The first data protection configuration or the second data protection configuration provides a partial integrity mode of data protection for frames of image data transmitted over the first data communication link. The partial integrity mode protects some, but not all frames of image data transmitted over the first data communication link using a message authentication code.

Abstract: Systems and techniques are described for securely managing maneuver identifications (IDs) for vehicle-based communications. For example, a method may include transmitting a message associated with a potential maneuver by the vehicle, determining a maneuver identification (ID) for the potential maneuver based on the message, and transmitting a first maneuver message (including the maneuver ID) for the potential maneuver. In another example, a method may include receiving a first maneuver message including a maneuver ID, determining that the first maneuver ID is a new maneuver ID, and determining, based on the first maneuver ID being a new maneuver ID, whether the first maneuver ID matches a second maneuver ID determined based on a previously-received message. In another example, a method may include receiving maneuver messages that include a same maneuver identification (ID) and one or more source IDs and determining whether the source ID(s) of the maneuver messages match.

Abstract: Systems and techniques media for image processing are described herein. For example, a device can obtain encryption mode information associated with an image. The encryption mode information includes a region of interest (ROI) of the image and/or a task associated with the image. The device can encrypt at least a portion of image based on the encryption mode information to produce an encrypted image. The device can transmit the encrypted image.

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment (UE) may detect a misbehavior condition in a local vehicle-to-everything (V2X) information source. The UE may receive an incoming V2X message. The UE may generate a misbehavior decision for the incoming V2X message based at least in part on the misbehavior condition in the local V2X information source. Numerous other aspects are described.

Abstract: A method of wireless communication at a first V2X device is described herein. The method includes obtaining a message including a MAC header, an L1 source ID, and an L1 destination ID. The method includes decoding the MAC header to produce an L2 source ID associated with the message based on at least one of (1) the L1 destination ID of the message matching a first number of bits of an L2 source ID of the first V2X device or (2) the L1 source ID of the message matching a second number of bits of the L2 source ID of the first V2X device. The method includes outputting an indication of a MAC cloning attack based on the L2 source ID of the message matching the L2 source ID of the first V2X device.

Abstract: Systems and techniques are described for object detection. For example, a computing device can apply a transformation to an image of a scene to generate a transformed image. The computing device can determine a plurality of candidate bounding regions for the transformed image. Each candidate bounding region is associated with an object in the scene. The computing device can determine a subset of candidate bounding regions for the transformed image by removing, using a non-max suppression model, at least one candidate bounding region of the plurality of candidate bounding regions. The computing device can generate an output bounding box for the object based on the subset of candidate bounding regions. The computing device can output the output bounding box. In some cases, the computing device can use an output of an image processing operation on the image to reduce a number of the plurality of candidate bounding regions.

Abstract: Various embodiments include methods and systems for vehicle-to-everything (V2X) system of a vehicle for managing processing of a basic safety message (BSM). Various embodiments may include calculating a relevancy value of a received BSM prior to conducting security processing of the BSM, determining whether the relevancy value of the BSM meets a relevancy threshold, and stopping processing of the BSM in response to determining that the relevancy value of the BSM does not meet the relevancy threshold.

Abstract: Disclosed are systems, apparatuses, processes, and computer-readable media for wireless communications. For example, a process may include determining a potential position overlap between a first vehicle and a second vehicle and determining a characteristic of at least one of the first vehicle or the second vehicle based on information from a vehicle-based message. The process may include determining whether the potential position overlap is an actual position overlap between the first vehicle and the second vehicle based on the characteristic.

Abstract: A user equipment (UE) for vehicle-to-everything (V2X) transmission is provided. The UE includes: at least one transceiver; at least one memory; and at least one processor, coupled with the at least one transceiver and the at least one memory, the at least one processor configured to: detect a second position of a second UE with respect to a first position of the UE; generate an alert time window based on the first position of the UE and the second position of the second UE; and schedule at least one transmission from the UE to the second UE based on the alert time window.