Abstract: Techniques and systems are provided for image processing. For instance, a process can include generating a set of surface representation values for one or more surfaces visible in one or more first images using a first three dimensional scene reconstruction (3DR) algorithm; preprocessing the set of surface representation values to generate preprocessing information for a second 3DR algorithm; generating, by the second 3DR algorithm, a refined set of surface representation values based on the set of surface representation values and the preprocessing information; and outputting the refined set of surface representation values.

Abstract: A radar module determines a two-dimensional velocity vector of a heavy-duty vehicle with respect to a ground plane supporting the vehicle. The system has a radar transceiver arranged to transmit and to receive a radar signal, via an antenna array, wherein the antenna array is configured to emit the radar signal in a first direction and in a second direction different from the first direction. The radar module has a processing device to detect first and second radar signal components of the received radar signal based on their respective angle of arrival, AoA, where the first radar signal component has an AoA corresponding to the first direction and the second radar signal component has an AoA corresponding to the second direction. The processing device determines the two-dimensional velocity vector of the heavy-duty vehicle based on respective Doppler frequencies of the first and second radar signal components.

Abstract: Disclosed are systems and techniques for three-dimensional reconstruction (3DR) of a scene. For example, a computing device can extract two-dimensional (2D) features from image frames of a scene. Each of the image frames includes a respective view of the scene. The computing device can unproject the 2D features from a 2D space onto a three-dimensional (3D) space to obtain 3D features. The computing device can determine weights for the 3D features based on geometric information and visual similarity information for the image frames. Each of the weights is associated with a respective feature 3D of the 3D features. The computing device can determine voxel features for the image frames based on the weights for the 3D features. The computing device can further form the 3DR of the scene based on the voxel features.

Abstract: A system and method are disclosed that leverage multi-factor authentication features of a service provider and intelligent call routing to increase security and efficiency at a customer call center. Pre-authentication of customer support requests reduces the potential for misappropriation of sensitive customer data during call handling. A contactless card uniquely associated with a client may provide a second factor of authentication to reduce the potential for malicious third-party impersonation of the client. Pre-authorized customer support calls are intelligently and efficiently routed in a manner that reduces the opportunity for malicious call interference and information theft.

Abstract: A processor-implemented method includes receiving neural network weights from an artificial neural network. The method also includes jointly embedding, by an encoder neural network, a first watermark into the neural network weights of the artificial neural network and a second watermark into an output of the artificial neural network to generate watermarked weights and watermarked output. The method further includes transmitting the watermarked weights and watermarked output. A processor-implemented method by a decoder neural network includes receiving watermarked neural network weights for an artificial neural network. The method also includes receiving a first private key. The method further includes decoding the watermarked neural network weights based on the first private key to obtain a first watermark.

Abstract: A receiver device includes detection logic, error counter logic, and threshold logic. The detection detects frame errors in data frames received by a transmitter device. The error counter logic increments a first value of an error count responsive to each error signal, indicative of a frame error in a data frame, received from the detection logic. The error counter logic reduces the first value to a second value (non-zero value) for the error count responsive to receiving a decrement signal and a period marker signal corresponding to a programmable period. The error counter logic resets the first value or the second value of the error count to zero responsive to receiving a reset signal. The threshold logic compares a current value of the error count with a threshold number of frame errors and output an interrupt responsive to the current value satisfying the threshold number of frame errors.

Abstract: Techniques for register renaming caching are described. In an embodiment, an apparatus includes a register renaming cache, front-end circuitry, lookup circuitry, and execution circuitry. The register renaming cache is to store register renaming information associated with an instruction trace. The register renaming information is to be learned from a first execution of the instruction trace and is to be used to perform register renaming in connection with a second execution of the instruction trace. The front-end circuitry is to provide, based on the instruction trace, operations for execution. The lookup circuitry to look in the register renaming cache for entries corresponding to the operations. The execution circuitry is to execute the instruction trace.

Abstract: A releasable restraint for an evacuation system includes a plug body and a socket body. A ball, a plunger, a first spring member, and an adjustable fastener are disposed at least partially within the plug body. A spring force of the first spring member is adjustable in response to moving the adjustable fastener with respect to the plug body. The first spring member urges the plunger against the ball, and in response, the ball extends through a sidewall of the plug body and at least partially through a sidewall of the socket body to lock the plug body to the socket body. In response to a tensile force applied to the releasable restraint, the ball is configured to retract at least partially into the plug body, against the urging of the first spring member, to release the socket body from the plug body.

Abstract: A vehicle motion management, VMM, system for a heavy-duty vehicle has at least one wheel speed sensor to output a wheel speed signal indicative of a rotation speed of a wheel on the vehicle, and at least one torque-generating device to apply torque to the wheel. The VMM system comprises processing circuitry arranged to determine at least the direction of a current applied torque at the wheel, and in preparation for determining wheel speed by the wheel speed signal, to apply a transient amount of torque to the wheel by the torque-generating device during a limited time period in a direction opposite to the direction of the current applied torque at the wheel.

Abstract: Devices, transistor structures, systems, and techniques are described herein related to selective front and backside contacts for stacked transistor devices. A transistor structure includes stacked first and second semiconductor structures with stacked first and second conductivity type source and drain structures coupled to the first and second semiconductor structures, respectively. A selective metal is on the frontside of first conductivity type source and a different metal is on the backside of the second conductivity type source. A deep via optionally having yet a different metal couples the frontside contact to backside metallization over the backside contact.

Abstract: A motion support device, MSD, control unit for a heavy duty vehicle, configured to control one or more MSDs associated with a wheel on the vehicle, wherein the MSD control unit is configured to be communicatively coupled to a vehicle motion management, VMM, unit for receiving control commands from the VMM unit comprising wheel speed and/or wheel slip requests to control vehicle motion by the one or more MSDs. The MSD control unit is configured to obtain a capability range indicating a range of wheel behaviors of the wheel for which the VMM unit is allowed to influence the behavior of the wheel by the control commands, monitor wheel behavior and to detect if wheel behavior is outside of the capability range, and trigger a control intervention function in case the monitored wheel behavior is outside of the capability range.

Abstract: A vehicle motion management (VMM) system for a heavy-duty vehicle, configured to obtain a desired wheel force value of a wheel of the vehicle; determine a torque limit for a first motion support device (MSD) associated with the wheel based on the desired wheel force value; determine a tire model based on a relationship between wheel force and wheel speed of the wheel; determine a desired wheel speed for the first MSD based on the tire model; and determine a torque fill request for a second MSD of the vehicle based on the desired wheel force and on a torque capability of the first MSD. The VMM system determines the torque fill request for the second MSD in dependence of the torque limit for the first MSD in case the operating torque of the first MSD is limited by the torque limit, and to determine the torque fill request for the second MSD in dependence of an applied torque status signal in case the operating torque of the first MSD is not limited by the torque limit.

Abstract: A radar module determines a two-dimensional velocity vector of a heavy-duty vehicle with respect to a ground plane supporting the vehicle. The system has a radar transceiver arranged to transmit and to receive a radar signal, via an antenna array, wherein the antenna array is configured to emit the radar signal in a first azimuth direction and in a second azimuth direction different from the first azimuth direction. The radar module further has a processing device arranged to detect first and second radar signal components of the received radar signal based on their respective angle of arrival, AoA, where the first radar signal component has an AoA corresponding to the first azimuth direction and the second radar signal component has an AoA corresponding to the second azimuth direction. The processing device is arranged to determine the two-dimensional velocity vector of the heavy-duty vehicle based on respective Doppler frequencies of the first and second radar signal components.

Abstract: A device includes a memory configured to store a first model and a second model. The first model is configured to perform inference based on a first set of parameters corresponding to a first context. The device includes one or more processors configured to process, using the second model, the first set of parameters and input corresponding to a second context to generate an output of the second model. The one or more processors are also configured to update the first model to perform inference using an updated set of parameters based on the output of the second model.

Abstract: An exemplary modular hold-open device is configured for use with a door closer comprising a body, a pinion rotatably mounted to the body, and an armature connected with the pinion. The modular hold-open device is configured to be mounted to the door closer, to selectively prevent rotation of the pinion by exerting on the pinion a resistive torque in a door-opening direction, and to cease exerting the resistive torque in response to a door-closing torque on the pinion exceeding a threshold torque to thereby permit rotation of the pinion in the door-closing direction.

Abstract: A retrofit module configured for use with a door closer having a pinion. The retrofit module generally includes a case, an output shaft, a motor, and a control assembly. The output shaft is rotatably mounted in the case, and is configured for rotational coupling with the pinion. The motor is mounted to the case, and is operable to rotate the output shaft in a door-opening direction. The control assembly is mounted to the case, and is configured to operate the motor to drive the output shaft in the door-opening direction in response to an actuating signal.

Abstract: A vehicle motion management, VMM, system for a heavy-duty vehicle has at least one wheel speed sensor to output a wheel speed signal indicative of a rotation speed of a respective wheel on the vehicle, at least one inertial measurement unit, IMU, to output an IMU signal indicative of an acceleration of the vehicle, a motion estimation function to estimate a vehicle motion state comprising vehicle speed over ground, based at least in part on the wheel speed signal and at least in part on the IMU signal, and a motion support device, MSD, coordination function to coordinate actuation of a plurality of MSDs of the heavy-duty vehicle in dependence of a vehicle motion request and the vehicle motion state. The motion estimation function models an error in the estimated vehicle motion state, and to output a free-rolling request to the MSD coordination function in case the modelled error fails to meet an acceptance criterion.

Abstract: A vehicle motion management, VMM, system for a heavy-duty vehicle has at least one radar module configured to determine a speed over ground of the heavy-duty vehicle, a motion estimation function configured to estimate a vehicle motion state based at least in part on the speed over ground, and a motion support device, MSD, coordination function configured to coordinate actuation of a plurality of MSDs of the heavy-duty vehicle in dependence of a vehicle motion request and the vehicle motion state. The radar module outputs a radar performance metric to the MSD coordination function indicative of an accuracy of the determined speed over ground. The MSD coordination function reduces a wheel slip set-point of one or more wheels of the heavy-duty vehicle in case the radar performance metric does not meet a pre-determined acceptance criterion.

Abstract: A computer-implemented method and apparatus for unsupervised anomaly detection is provided. The method includes identifying one or more unsupervised anomaly detection approaches, wherein cach anomaly detection approach identified includes an anomaly detection algorithm and a corresponding threshold parameter value; and receiving a set of data. The method further includes, for the identified anomaly detection approaches, applying a statistical method including: sampling over the identified anomaly detection approaches to obtain a prior probability distribution; obtaining an input of anomalies and non-anomalies for at least a portion of the received set of data; obtaining a post probability distribution over the identified anomaly detection approaches based on the obtained input, wherein the post probability distribution updates the prior probability distribution; and determining whether a first stopping criterion is met and, if the first stopping criterion is not met, reapplying the statistical method.

Abstract: A vehicle motion management, VMM, system for a heavy-duty vehicle has at least one wheel speed sensor outputs a wheel speed signal indicative of a rotation speed of a respective wheel on the vehicle, at least one inertial measurement unit, IMU, outputs an IMU signal indicative of an acceleration of the vehicle, a motion estimation function estimates a vehicle motion state comprising vehicle speed over ground, SOG, based on the at least one wheel speed signal and on the at least one IMU signal. The motion estimation function estimates a respective SOG error associated with the wheel speed signal as an increasing function of an applied torque to the wheel, and a respective SOG error associated with the IMU signal as an increasing function of a time duration elapsed after calibration of an integrator of the IMU signal.