Abstract: Embodiments of the present disclosure provide a centralized light supply apparatus, including: a controller, at least one light source pool, and a plurality of CPO switches. The controller is connected with the light source pool and the plurality of CPO switches through a two-way communication channel, and the light source pool is connected with the plurality of CPO switches through a photoelectric hybrid connector. The controller is configured to control the light source pool and the plurality of CPO switches. The light source pool is configured to output a light source to the plurality of CPO switches under control of the controller. The plurality of CPO switches are configured to modulate the received light source into an optical signal under control of the controller and output the optical signal. The light source pool may be cascaded and expanded through one controller.

Abstract: In one embodiment, a method includes accessing video captured by one or more cameras of a head-mounted device (HMD) worn by a user and determining, from the accessed video, multiple environmental feature maps and multiple corresponding environmental embeddings representing an environment of the user captured in the accessed video. The method further includes fusing, by multiple trained environmental transformer models, the environmental embeddings to create a fused environmental embedding; determining a head pose of the user coincident with the accessed video; and predicting, by a trained decoder and based on the fused environmental embedding and the determined head pose of the user, a body pose of the user.

Abstract: An abrasive block mounting and dismounting assembly, a mounting and dismounting method, a wheel tread cleaning device, and a railway vehicle.

Abstract: An interactive sexual device is disclosed. The interactive sexual device has a fixation portion configured to be secured within a first orifice of a wearer, and a sensing module, which is configured to monitor interaction information of a sexual interaction between the wearer and a partner of a wearer, is configured to be located approximately toward a body of the partner during the sexual interaction when the fixation portion is secured within the first orifice, the interaction information including at least one of physical contact information or distance information between the sensing module and the partner. The sexual interaction includes the partner's penis being received within a second orifice of the wearer, which is adjacent to the first orifice, during intercourse between the wearer and the partner.

Abstract: A method for preparing a graphene film with a micrometer groove array and a triangular cross-section includes the following steps: A, subjecting a flexible sheet material to a pre-treatment, and obtaining a flexible sheet substrate; B, coating a carbon precursor material onto the flexible sheet substrate according to a required thickness of a solid carbon precursor, and conducting curing to obtain a flexible composite film loaded with the solid carbon precursor; where the carbon precursor material has a triazine ring network cross-linking structure; and C, adjusting laser parameters, and conducting direct-writing scanning on a surface of the solid carbon precursor with laser, followed by cleaning and drying to obtain the graphene film with a micrometer groove array and a triangular cross-section. The method solves the problems of current traditional processing techniques, which struggle to meet the high precision and three-dimensional requirements of a micro-scale/nano-scale structure of graphene films.

Abstract: An electronic device may be provided to perform collision analysis. The electronic device may be configured to: receive sensor data from an inertial-measuring-unit (IMU) sensor; detect that a collision event has occurred based on the sensor data and detect collision data from the sensor data; input the collision data into a collision classification model that is trained to predict a collision material type and a collision intensity of the collision event; and determine an action to be performed by the electronic device based on the collision material type and the collision intensity of the collision event.

Abstract: A method of performing visual question answering (VQA), including: obtaining an image and a question corresponding to the image; generating a plurality of feature predictions about at least one object included in the image by providing the image to an artificial intelligence (AI) scene perception model; generating a plurality of symbolic programs and a plurality of program confidence scores by providing the question to an AI question parsing model; selecting a symbolic program associated with a program confidence score which is highest among the plurality of program confidence scores; executing the selected symbolic program by performing a set of logic operations included in the selected symbolic program on the plurality of feature predictions; and determining a natural language answer to the question based on a result of the set of logic operations.

Abstract: A flame-proof portable power supply includes a battery pack, an electrical assembly, a metal housing and an aerosol fire extinguishing device. A holding cavity for installing the battery pack or the electrical assembly are formed in the metal housing, the battery pack is electrically connected to the electrical assembly, and the battery pack is directly connected to the metal housing. The aerosol fire extinguishing device is disposed in a holding cavity, and the aerosol fire extinguishing device is configured to release aerosol in the holding cavity when temperature in the holding cavity reaches a specified temperature value.

Abstract: A method for detecting a drop event of an electronic device, may include: obtaining an angular velocity and a proper acceleration of the electronic device based on sensor data received from an inertial measurement unit (IMU) sensor of the electronic device; obtaining centripetal acceleration of the electronic device based on the angular velocity, principal moments of inertia of the electronic device, and position of the IMU sensor within the electronic device; based on an acceleration difference between the centripetal acceleration and the proper acceleration, determining whether the electronic device is in a fall state; and based on the electronic device being determined to be in the fall state, providing an analysis result of the drop event.

Abstract: The present application discloses a two-phase sulfate reduction device and treatment method for preventing sludge calcification, which belongs to the technical field of wastewater treatment in environmental engineering. The present application combines a two-phase sulfate reduction anaerobic reactor with an alloy catalyst, and reasonably arranges a two-phase connection unit of the alloy catalyst within the device body. By using the special functions of the porous alloy catalyst material to release free electrons into a body of water and polarize the body of water, the electrostatic potential of the body of water is changed, sludge calcification during the wastewater treatment process is prevented, and the treatment efficiency of the anaerobic system for sulfate organic wastewater is ensured.

Abstract: An end cover assembly, an air cylinder, a tread sweeper and a railway vehicle.

Abstract: A method includes receiving, at an electronic device and from a second electronic device, a second root identifier, wherein the second root identifier is associated with a second profile tree maintained at the second electronic device and determining that a first root identifier does not match the second root identifier, wherein the first root identifier is based on a first profile tree maintained at the electronic device. The method further includes sending, to the second electronic device, the first profile tree, wherein the first profile tree comprises representation of currently maintained user profiles at the electronic device, receiving, from the second electronic device, user profile update information, and updating a subset of the currently maintained user profiles based on the user profile update information.

Abstract: In one embodiment, a method includes accessing a plurality of initial gradients associated with a machine-learning model from a data store associated with a first electronic device, selecting one or more of the plurality of initial gradients for perturbation, generating one or more perturbed gradients for the one or more selected initial gradients based on a gradient-perturbation model, respectively, wherein for each selected initial gradient: an input to the gradient-perturbation model comprises the selected initial gradient having a value x, the gradient-perturbation model changes x into a first continuous value with a first probability or a second continuous value with a second probability, and the first and second probabilities are determined based on x, and sending the one or more perturbed gradients from the first electronic device to a second electronic device.

Abstract: The present invention discloses methods and platforms for generating protein binding proteins with specificity for native peptide-MHC (pMHC) complexes. The pMHC binding proteins can be used in bi-specific antibodies or for generating CAR T cells capable of binding to peptides bound to specific MHC alleles.

Abstract: The present disclosure relates to a multi-modal imaging device. The multi-modal imaging device includes: a Raman spectroscopic analysis module configured to obtain Raman spectroscopic information of a target object on a first sampling position by using excitation light; an optical coherence tomography module configured to obtain a tissue structure image of the target object on a second sampling position by using imaging detection light; and a co-localization module configured to control the first sampling position of the excitation light in the Raman spectroscopic analysis module and/or the second sampling position in the optical coherence tomography module according to a determined concerned area of the target object, so that the first sampling position and the second sampling position are spatially co-localized in the concerned area.

Abstract: Methods, systems, and apparatuses for performing natural language prompt compression, the method being performed by an electronic device and including: obtaining a prompt for an artificial intelligence (AI) inference model, wherein the prompt corresponds to a first plurality of tokens; providing the prompt as an input to an AI compression model; and obtaining a compressed prompt based on an output of the AI compression model, wherein the compressed prompt corresponds to a second plurality of tokens which is smaller than the first plurality of tokens, and wherein the prompt and the compressed prompt are expressed using natural language.

Abstract: In one embodiment, a method includes accessing video captured by one or more cameras of a head-mounted device (HMD) worn by a user and determining, from the accessed video, multiple environmental feature maps and multiple corresponding environmental embeddings representing an environment of the user captured in the accessed video. The method further includes fusing, by multiple trained environmental transformer models, the environmental embeddings to create a fused environmental embedding; determining a head pose of the user coincident with the accessed video; and predicting, by a trained decoder and based on the fused environmental embedding and the determined head pose of the user, a body pose of the user.

Abstract: The present disclosure belongs to the technical field of drug delivery and biomimetic structures inducing cell proliferation and differentiation, and in particular relates to a micro-nano medical stent for directional, quantitative and timed controlled release of drug and a preparation method thereof. The micro-nano medical stent provided by the present disclosure is composed of an ordered fiber structure electrospun as a skeleton layer and a disordered fiber structure electrospun as an ECM layer that can both load drugs. The micro-nano medical stent can achieve directional, quantitative and timed release of drugs.

Abstract: A method includes receiving wireless fingerprint data identifying multiple locations within a specified area and, for each location, one or more signal strength values associated with wireless signals received from one or more of multiple wireless transmitters. The wireless fingerprint data is missing signal strength values for one or more transmitters at one or more specific locations. The method also includes generating a training dataset by adding filler signal strength values in place of at least some missing values. The method further includes training a machine learning model using the training dataset. The model is trained to receive a specified location as input and generate predicted signal strength values as outputs. In addition, the method includes using the trained model to generate additional signal strength values. At least some additional signal strength values are to be used in place of at least a portion of the missing values.