Abstract: Apparatuses, systems, and techniques to cause a first tensor to be translated into a second tensor according to a tensor map. In at least one embodiment, one or more circuits are to perform an application programming interface (API) to cause a first tensor to be translated into a second tensor according to a tensor map.

Abstract: A control method for cooling a refrigerant of an inverter includes the steps of: sucking a portion of the refrigerant out from a reservoir of a condenser in a chiller into a cooling plate of the inverter for cooling; according to a pressure value of the condenser and an outlet pressure value of the inverter, determining a set condition; and, based on the set condition, controlling an opening of an electronic expansion valve to adjust a flow of the portion of the refrigerant in the cooling plate of the inverter. In addition, a refrigerant bypass pipeline of the inverter is also provided.

Abstract: Two-dimensional face presentation attacks are one of most notorious and pervasive face spoofing types, causing security issues to facial authentication systems. To tackle these issues, a cost-effective face anti-spoofing (FAS) system based on acoustic modality, named as Echo-FAS, is devised, which employs a crafted acoustic signal to probe the presented face. First, a large-scale, high-diversity, acoustic-based FAS database, named as Echo-Spoof, is built. Based upon Echo-Spoof, we design a two-branch framework combining global and local frequency features of the presented face to distinguish live vs. spoofing faces. Echo-FAS has the following merits: (1) it only needs one speaker and one microphone; (2) it can capture three-dimensional geometrical information of the presented face and achieve a remarkable FAS performance; and (3) it can be handily allied with RGB-based FAS models to mitigate the overfitting problem in the RGB modality and make the FAS model more accurate and robust.

Abstract: The present disclosure provides an ultrasonic imaging method and apparatus. The apparatus includes an ultrasonic signal emission source and a receiving array including ultrasonic signal receiving circuits; the method includes: turning on the source to perform emission of ultrasonic waves toward an object to be detected and causing the ultrasonic waves to propagate through the object, a depth-wise direction of which is along a propagation direction of the ultrasonic waves; after a first predetermined time period having elapsed since the source was turned on, turning on the receiving array to receive reflected echoes returning from a first section plane of the object to be detected perpendicular to the depth-wise direction; thereafter, turning off the receiving array, storing the reflected echoes by the ultrasonic signal receiving circuits and acquiring reflected echo signals, and successively reading the reflected echo signals from the ultrasonic signal receiving circuits during a reading time period.

Abstract: A machine learning attack resistant strong PUF with a dual-edge sampling function comprises switch units, a first arbiter and a second arbiter. The first arbiter is for determining a sequential order of delays at a rising edge of signals input to a first input terminal and a second input terminal of the first arbiter. The second arbiter is for determining a sequential order of delays at a falling edge of signals input to a first input terminal and a second input terminal of the second arbiter. Each switch unit is composed of eight MOS transistors. The strong PUF has a high capacity to resist machine learning attacks and small hardware expenditure through simple structural design of the switch units, realizing machine learning attack resistance and small hardware expenditure at the same time, and generating a large number of challenge response pairs through dual-edge sampling realized by the two arbiters.

Abstract: A distributed aperture radar system, apparatus, architecture, and method are provided for generating a completed virtual array aperture by using a radar control processing unit to construct a sparse forward-backward matrix from one or more sparse measurement array vectors and to generate beamforming outputs of a completed virtual array aperture by performing forward-backward matrix completion processing of the sparse forward-backward matrix to construct a completed forward-backward matrix by filling holes in the sparse MIMO virtual array aperture to suppress spurious sidelobes caused by holes in the sparse forward-backward matrix, thereby enabling computation of one or more super resolution direction of arrival (DoA) estimation values based on the completed forward-backward matrix.

Abstract: In various embodiments, a scene reconstruction model generates three-dimensional (3D) representations of scenes. The scene reconstruction model computes a first 3D feature grid based on a set of red, blue, green, and depth (RGBD) images associated with a first scene. The scene reconstruction model maps the first 3D feature grid to a first 3D representation of the first scene. The scene reconstruction model computes a first reconstruction loss based on the first 3D representation and the set of RGBD images. The scene reconstruction model modifies at least one of the first 3D feature grid, a first pre-trained geometry decoder, or a first pre-trained texture decoder based on the first reconstruction loss to generate a second 3D representation of the first scene.

Abstract: A method for modeling a subsurface property for a subterranean volume of interest includes receiving input measurement data representing a subterranean volume of interest, predicting a subsurface property based at least in part on the input measurement data using a first machine learning model, predicting a subsurface property model based at least in part on the subsurface property, the input measurement data, or both, using a second machine learning model, predicting synthetic measurement data based at least in part on the subsurface property model using a third machine learning model, a physics-based model, or both, comparing the synthetic measurement data and the input measurement data, and training the first machine learning model, the second machine learning model, or both based at least in part on the comparing.

Abstract: Aspects of the invention include techniques for intimacy-based masking of three dimensional (3D) face landmarks within a metaverse. A non-limiting example method includes determining a degree of intimacy R between a first user and a second user responsive to a first avatar of the first user being observed by a second avatar of the second user in a virtual environment. An initial set of M landmarks of a feature of the first avatar is determined and N landmarks are randomly selected from M for masking. The N landmarks are replaced with random noise sampled with a unit normal distribution to define a set of noised landmarks X that are then iteratively denoised to define a set of denoised landmarks D. A masked version of the first avatar is presented to the second user by modifying the feature of the first avatar with the set of denoised landmarks D.

Abstract: The present disclosure relates to a display device with an outer contour of a polygon. The display device includes: a display panel; a functional panel on a side of a light exiting surface of the display panel; a first circuit board electrically connected to the display panel and bent at a first edge of the polygon to a side of the display panel away from the light exiting surface of the display panel; and a second circuit board electrically connected to the functional panel and bent at a second edge of the polygon to the side of the display panel away from the light exiting surface of the display panel, wherein the first edge and the second edge are different edges, and an orthographic projection of the first circuit board on the light exiting surface of the display panel does not overlap with an orthographic projection of the second circuit board on the light exiting surface of the display panel.

Abstract: Apparatuses, systems, and techniques to create one or more memory transaction software objects. In at least one embodiment, one or more circuits are to perform an application programming interface (API) to cause one or more software objects to indicate whether one or more memory transactions have been performed.

Abstract: The invention belongs to the technical field of biomedical testing, and particularly relates to a biomarker for diagnosis, treatment, and prognosis for hepatocellular carcinoma bone metastasis (HCC) and application thereof. The application of biomarker VAPA for preparing a reagent or a kit for diagnosis, treatment, and prognosis for HCC bone metastasis. The application of the reagent that detects VAPA expression levels in blood and tissues for preparing a reagent or kit for diagnosis, treatment, and prognosis for HCC bone metastasis. Compared with the diagnostic kits in the prior art that can only detect bone metastasis after it has occurred, the VAPA described in this invention offers a more specific and sensitive assay kit for the early prediction, diagnosis, disease progression assessment, treatment efficacy evaluation, drug guidance, and prognosis assessment of HCC bone metastasis. Additionally, the VAPA may also serve as a target for therapeutic interventions in bone metastasis.

Abstract: Apparatuses, systems, and techniques to cause information to be provided. In at least one embodiment, one or more circuits are to perform an application programming interface (API) to cause an amount of information to be accessed as a result of one or more memory transactions to be provided to one or more users.

Abstract: A computer-implemented method is provided.

Abstract: In some implementations, a memory device may determine that a power loss has occurred. The memory device may determine a last written page (LWP) location associated with an LWP of a block of a memory of the memory device. The memory device may determine one of: a word line group (WLG) associated with the LWP location and at least one WLG-dependent offset associated with the WLG, or a partial block (PB) fill ratio associated with the LWP location and at least one PB-fill-ratio-dependent offset associated with the PB fill ratio. The memory device may perform a power loss error detection procedure based on one of the at least one WLG-dependent offset or the at least one PB-fill-ratio offset by applying the one of the at least one WLG-dependent offset or the at least one PB-fill-ratio offset to at least one read reference voltage.

Abstract: A prediction-based system and method for trajectory planning of autonomous vehicles is configured to: receive data from a training data collection system, the training data including perception data and context data corresponding to human driving behaviors; perform a training phase for training a trajectory prediction module using the training data; receive perception data associated with a host vehicle; and perform an operational phase for extracting host vehicle feature data and proximate vehicle context data from the perception data, generating a proposed trajectory for the host vehicle, using the trained trajectory prediction module to generate predicted trajectories for each of one or more proximate vehicles near the host vehicle based on the proposed host vehicle trajectory, determining if the proposed trajectory for the host vehicle will conflict with any of the predicted trajectories of the proximate vehicles, and modifying the proposed trajectory for the host vehicle until conflicts are eliminated.

Abstract: The present disclosure provides a method for anti-tampering apparatus servicing data implemented by a calculation device connected to a target device, the method comprising: identifying a contract identification code and obtaining a contract package file and a contract authentication code from at least one remote device; obtaining a microservice file corresponding to the target device from the remote device when a device embedded code of the calculation device is matching the contract authentication code; performing the microservice file to enable the target device according to the contract package file and generate an execution report; publishing the execution report to the remote device to obtain an acceptance certification code; and combining and hashing the device embedded code, the contract authentication code and the acceptance certification code to generate a hash value, and sending the hash value to a blockchain.

Abstract: The present disclosure discloses a universal method for controlling recharge of robot, chip and robot, the method includes: S1, whether the robot has detected the navigation target point or not is determined, if yes, the robot is controlled to move along a planned navigation path, otherwise, the process proceeds to S3; in the moving process, when the charging station signal is received or all the pre-marked navigation target points are traversed, the process proceeds to S2; S2, whether the robot has received the charging station signal or not is determined, if yes, charging on the station is carried out, otherwise, the charging station signal is searched, and when the robot does not receive the charging station signal, the process returns to S1; S3: the robot is controlled to search for the charging station signal until searching the charging station signal, returning to S2, otherwise, walking along the boundary is continued.

Abstract: Apparatuses, systems, and techniques to generate one or more neural networks. In at least one embodiment, one or more neural networks are generated, based on, for example, one or more convolutional neural network operations and one or more transformer neural network operations.

Abstract: A cart that reduces impact on a load placed on a loading platform 50. The cart 10 includes a drive that raises/lowers the platform 50 by rotating a motor which extends/retracts a mechanism 70 vertically. The drive rotates the motor when an external force raises/lowers the platform 50 and extends/retracts the mechanism 70. When a control device detects lowering of the platform 50 held at a predetermined vertical position and detects an increase in weight of the load on the platform 50 based on an output signal of a rotation sensor that detects rotation of the motor and an output signal of a current sensor that detects a drive current of the motor, the control device controls the motor such that the platform 50 is raised and held at the either the predetermined vertical position or a vertical position lower than that position.