Abstract: In a method for controlling an associated virtual object, a virtual scene of a virtual environment including a first virtual object is output for display. The first virtual object is controlled to move in the virtual environment. A state of the associated virtual object of the first virtual object is controlled to switch between a first state and a second state based on a first user input. The associated virtual object is output for display based on the state of the associated virtual object. In the first state, the associated virtual object is carried by the first virtual object and configured to perform a first function. In the second state, the associated virtual object is configured to move separately from the first virtual object and perform a second function that is different from the first function.

Abstract: A method for dynamic modeling and manipulation of multi-object scenes is described. The method includes using object-centric neural implicit scattering functions (OSFs) as object representations in a model-predictive control (MPC) framework for the multi-object scenes. The method also includes modeling a per-object light transport to enable compositional scene re-rendering under object rearrangement and varying lighting conditions. The method further includes applying inverse parameter estimation and graph neural network (GNN) dynamics models to estimate initial object poses and a light position in the multi-object scene. The method also includes manipulating an object perceived in the multi-object scene according to the applying of the inverse parameter estimation and the GNN dynamics models.

Abstract: The present disclosure provides techniques for robot policy training from view-invariant demonstrations of a task. An example method includes obtaining an image of an environment of the apparatus; generating a plurality of random pose transforms to apply to the image; generating, with a generative diffusion model, respective augmented images of the image based on each of the plurality of random pose transforms, wherein the respective augmented images correspond to augmented views of the environment; selecting a set of the respective augmented images based on a distribution corresponding to a sphere centered at the robot base; and training a robot task diffusion policy with the set of the respective augmented images.

Abstract: An electric sofa anti-pinch system and an electric sofa are provided. The electric sofa anti-pinch system includes a thin film pressure sensor, a drive module, and a control module. A triggering mechanism of the thin film pressure sensor is provided in a squeezing risk area of a moving part in the electric sofa; the thin film pressure sensor is configured to output a detection signal when the triggering mechanism is squeezed; the drive module is connected to the moving part and configured to drive the moving part to operate; the control module is respectively connected to the thin film pressure sensor and the drive module and configured to control the moving part to stop operating or reversely operate through the drive module when receiving the detection signal. Embodiments of the present disclosure can improve a reliability of the electric sofa anti-pinch system, enhance an anti-pinch effect.

Abstract: A fluid flow system for a rotatable shaft assembly includes a first fluid flow device installed in a rotatable shaft, and a second fluid flow device installed in the rotatable shaft, spaced axially apart from the first fluid flow device. The first fluid flow device includes a wall portion having a radial dimension that retains oil in the rotatable shaft. The second fluid flow device includes a wall portion having a radial dimension that facilitates discharge of oil from the radial shaft. The first and second fluid flow devices each include a plurality of baffles on the respective wall portions thereof. The plurality of baffles drive the flow of fluid in an axial direction, from the first fluid flow device toward the second fluid flow device, as the first and second fluid flow devices rotate together with the rotatable shaft.

Abstract: A method for determining a complexity of a natural language query includes converting a first natural language query into executable program code, the first natural language query being a query for a first video to be answered by one or more video question answering (VideoQA) models. The method also includes generating, via a complexity model, an abstract syntax tree (AST) based on the executable program code. The method further includes determining, via the complexity model, a complexity of the first natural language query based on quantity of subtrees, from a group of subtrees, that are present in the AST.

Abstract: One or more variable stiffness devices may be positioned between the splined coupling of a pair of rotating mechanical components to reduce or substantially eliminate backlash during operation. The variable stiffness device may include a spring pin that exerts a variable biasing force on mating surfaces of the pair of rotating mechanical components to maintain a relative position of the components in response to rotational and/or torsional forces exerted during operation.

Abstract: A power tool includes a housing, an electric motor, a switch assembly, and a directional assembly. The motor is disposed at least partially in the housing and includes a rotating shaft rotatable about a first axis. The switch assembly is operable to switch between a startup position and a shutdown position. The switch assembly sends a startup signal to the motor at the startup position. The switch assembly sends a shutdown signal to the motor at the shutdown position. The directional assembly is for the user to operate to control a rotation direction of the electric motor. The directional assembly is further configured to mate with the switch assembly to lock the switch assembly at the startup position. When the directional assembly is operated to be disengaged from the switch assembly to unlock the switch assembly, the switch assembly switches to the shutdown position.

Abstract: One or more variable stiffness devices may be positioned between the splined coupling of a pair of rotating mechanical components to reduce or substantially eliminate backlash during operation. The variable stiffness device may include a spring pin that exerts a variable biasing force on mating surfaces of the pair of rotating mechanical components to maintain a relative position of the components in response to rotational and/or torsional forces exerted during operation.

Abstract: Provided are a cinema chair control system and method, and a cinema chair. The cinema chair control system includes multiple near-end controllers configured on cinema chairs, a gateway, and a first remote control apparatus; the multiple near-end controllers are each configured with a communication module and a drive module, the communication modules may communicate with each other, and the drive module is configured to receive a control instruction sent from the communication module to control the cinema chair; the gateway is in a communication connection with the communication modules and is connected to the Ethernet; and the first remote control apparatus is configured to send the control instruction to at least one near-end controller through the gateway.

Abstract: Disclosed are a low-noise electromagnetic valve assembly, an air bag massage device, and a Control Method. The low-noise electromagnetic valve assembly includes an electromagnetic valve body, a flow guide piece, and a silencer; and the flow guide piece includes an air intake interface, an air release interface, an air bag interface, and a reserved interface The electromagnetic valve body enables the air bag interface to be in communication with the air intake interface or the air release interface, and the reserved interface is in communication with the air bag interface. The air release interface is in communication with the silencer, and the silencer is provided with a silencing air outlet. The low-noise electromagnetic valve assembly, the air bag massage device, and the control method can solve the problems of high production cost and poor usage experience of an electromagnetic valve.

Abstract: This application provides a data processing method and a related apparatus. The method includes: obtaining K-dimensional data, and inputting the K-dimensional data into a first machine learning model, to obtain a solution to a to-be-solved problem. The first machine learning model includes a first processing module and a second processing module. The second processing module is determined based on a constraint condition of the to-be-solved problem. The second processing module is configured to perform dimensional generalization on the K-dimensional data. The first processing module is obtained through training based on m-dimensional data. A value of m is irrelevant to a value of K, and K and m are positive integers.

Abstract: Disclosed herein are a virtual object control method and apparatus, a device, a medium, and a product, which relate to the field of virtual environments. This method includes: displaying a master virtual object and an interactive element in a virtual scene; controlling, in response to an attack operation on the interactive element, the master virtual object to attack the interactive element using a virtual prop; controlling, in response to an accessory switching operation, the master virtual object to switch a first prop accessory of the virtual prop to a second prop accessory; and detaching, after the first prop accessory is switched, the first prop accessory from the virtual prop, and transforming the first prop accessory into a functional prop.

Abstract: A fluid flow system for a rotatable shaft assembly includes a first fluid flow device installed in a rotatable shaft, and a second fluid flow device installed in the rotatable shaft, spaced axially apart from the first fluid flow device. The first fluid flow device includes a wall portion having a radial dimension that retains oil in the rotatable shaft. The second fluid flow device includes a wall portion having a radial dimension that facilitates discharge of oil from the radial shaft. The first and second fluid flow devices each include a plurality of baffles on the respective wall portions thereof. The plurality of baffles drive the flow of fluid in an axial direction, from the first fluid flow device toward the second fluid flow device, as the first and second fluid flow devices rotate together with the rotatable shaft.

Abstract: In a method for controlling a companion object, a virtual scene is displayed. The virtual scene includes a first virtual object and a companion object of the first virtual object. The companion object is configured to switch between a first state and a second state. The companion object is controlled to switch between the first state and the second state. When the companion object is in the first state, the companion object is in a first form that is attached to the first virtual object. When the companion object is in the second state, the companion object is (i) in a second form that is not attached to the first virtual object and (ii) configured to move separately from the first virtual object, the first form of the companion object being different from the second form of the companion object.

Abstract: An axle decoupler system comprises: an axle decoupler configured to selectively couple an axle of a vehicle with a wheel hub of the vehicle, the axle decoupler comprising: a sliding ring having first splines coupled to second splines on the axle, the sliding ring having a first position where the first splines are not also coupled to third splines on the wheel hub, and a second position where the first splines are also coupled to the third splines; and an electric motor for moving the sliding ring into the first or second position; and a one-way clutch coupling the axle and the wheel hub to each other.

Abstract: The present disclosure relates to neural network construction methods. One example method includes generating parameters of a target neural network based on a parameter generation network, where input of the parameter generation network includes information about a relative number of a neuron in the target neural network, the relative number of the neuron represents a relative location of the neuron at a first neural network layer, and the first neural network layer is a layer at which the neuron in the target neural network is located. The target neural network is constructed based on the parameters of the target neural network, where the parameters of the target neural network are generated by inputting the information about the relative number of the neuron into the target neural network.

Abstract: A method for dynamic modeling and manipulation of multi-object scenes is described. The method includes using object-centric neural implicit scattering functions (OSFs) as object representations in a model-predictive control (MPC) framework for the multi-object scenes. The method also includes modeling a per-object light transport to enable compositional scene re-rendering under object rearrangement and varying lighting conditions. The method further includes applying inverse parameter estimation and graph neural network (GNN) dynamics models to estimate initial object poses and a light position in the multi-object scene. The method also includes manipulating an object perceived in the multi-object scene according to the applying of the inverse parameter estimation and the GNN dynamics models.

Abstract: An axle decoupler system comprises: an axle decoupler configured to selectively couple an axle of a vehicle with a wheel hub of the vehicle, the axle decoupler comprising: a sliding ring having first splines coupled to second splines on the axle, the sliding ring having a first position where the first splines are not also coupled to third splines on the wheel hub, and a second position where the first splines are also coupled to the third splines; and an electric motor for moving the sliding ring into the first or second position; and a one-way clutch coupling the axle and the wheel hub to each other.

Abstract: A method for scene synthesis from human motion is described. The method includes computing three-dimensional (3D) human pose trajectories of human motion in a scene. The method also includes generating contact labels of unseen objects in the scene based on the computing of the 3D human pose trajectories. The method further includes estimating contact points between human body vertices of the 3D human pose trajectories and the contact labels of the unseen objects that are in contact with the human body vertices. The method also includes predicting object placements of the unseen objects in the scene based on the estimated contact points.