Abstract: The present disclosure is directed toward systems, methods, and non-transitory computer readable media that train, utilize, and/or evaluate multimodal large language models in visual text grounding including identifying grounding text content within digital images that supports text answers to text questions. For example, the disclosed systems receive a text question associated with a digital image that includes text content. Furthermore, the disclosed systems generate a large language model prompt which includes the digital image, the text question, and an instruction to generate a text answer and identify a portion of the text content of the digital image that supports the text answer. The disclosed systems utilize the large language model prompt to prompt a multimodal large language model to generate a text answer to the text question from the text content of the digital image and generate a text description that identifies grounding text content supporting the text answer.

Abstract: Disclosed in the present application are a surgical robotic arm and a surgical robot. In the surgical robotic arm provided by the present application, a first rotary arm is rotatably connected to a second rotary arm; a power unit is provided on the first rotary arm, and an execution unit is provided on the second rotary arm; the first rotary arm is provided with a first transmission assembly, and the second rotary arm is provided with a second transmission assembly; the driving force of the power unit is transmitted to the second transmission assembly by means of the first transmission assembly and a drive shaft, and then the second transmission assembly drives the execution unit to act; a driving axis of the execution unit is not parallel to a driving axis of the power unit.

Abstract: The present application discloses an encoding and decoding method, a point cloud encoder, a point cloud decoder, and a storage medium. At an encoding end, when a current child node enables an inter-frame prediction mode, a target entropy encoding parameter set of the current child node is determined on the basis of occupancy information of a predicted child node; a target entropy encoding parameter of the current child node is determined from the target entropy encoding parameter set on the basis of context information of the current child node; and entropy encoding is performed on the occupancy information of the current child node on the basis of the target entropy encoding parameter, and an obtained encoded bit is written into a bitstream. At a decoding end, the same method is used to determine a target entropy decoder, and entropy decoding is performed on occupancy information.

Abstract: A video decoding technique includes parsing a bitstream to determine a quantization parameter (QP) for a luma component of a region from a data unit of a parameter set included in the bitstream, determining a flag from the data unit of the parameter set, determining, in a case that the flag is equal to a first value, a QP for a chroma component of the region as a first function of the QP for the luma component and a default chroma delta QP indicated by the flag, determining, in a case that the flag is equal to a second value, the QP for the chroma component of the region by (a) obtaining a delta chroma QP from the data unit, and (b) determining the QP for the chroma component as a second function of the QP for the luma component and the delta chroma QP, and decoding the chroma component using the QP for the chroma component in a case that the parameter set is activated for decoding the region.

Abstract: A vehicle, having a vehicle door assembly. The vehicle door assembly includes a vehicle door, an unlocking assembly and a handle assembly. The vehicle door is provided with a handle groove. The unlocking assembly is arranged inside the vehicle door. The handle assembly is arranged in the handle groove. The handle assembly includes a handle and a first pivot shaft. The handle is provided with a retracted position and a position to be unlocked. When the handle is in the retracted position, the handle is located at an opening of the handle groove and does not extend beyond a surface of the vehicle door. When the handle is in the position to be unlocked, the handle is located in the handle groove, and the handle comes into contact with the unlocking assembly.

Abstract: A GAN-based point cloud processing method, decoder and computer-readable storage medium are provided. The method includes: a first point cloud is input into a feature extraction module to obtain a first feature of a point in the first point cloud; the first feature is input into a coarse feature extension module to obtain a second feature; the second feature is input into a geometry generation module to map the second feature from a feature space to a geometric space to obtain a second point cloud; the second feature and the second point cloud are input into a geometric offset prediction module to determine the offset of points in the second point cloud in the geometric space; and a third point cloud according to the second point cloud and the offset is obtained.

Abstract: A method, system and apparatus for performing a cell change procedure in a hybrid terrestrial network-non terrestrial network (TN-NTN) are disclosed. According to one aspect, a method in a user equipment (UE) includes determining a qualification status of at least a first cell of a first TN node and a second cell of a first NTN node. The method includes selecting a target cell for a cell change, the target cell being one of a qualified first cell of the first TN node and a qualified second cell of the first NTN node, the target cell being selected based at least in part on at least one of: a low activity multicarrier, LAM, measurement capability of the UE; a registration area of a candidate target cell; and a state of a global navigation satellite system, GNSS, receiver of the user equipment. The method also includes performing the cell change to the selected target cell.

Abstract: A backlight module for keyboard, comprising keycaps, upper conduction components, lower conduction components, LED lights, light shielding film, light guiding film and light reflecting film, the keycap is arranged above the upper conduction component, the lower conduction component is arranged below the upper conduction component, an opening is provided on the lower conduction component, the LED light is arranged under the lower conduction component, the light shielding film is arranged under the lower conduction component, and is provided with a first through hole corresponding to the LED light, the light guiding film is arranged under the light shielding film, which is provided with a second through hole corresponding to the LED light, which is arranged in the first and second through hole, the corresponding openings on the light guiding film are provided with light guiding dots, the light reflecting film is arranged under the light guiding film.

Abstract: The present disclosure relates to the technical field of optical fiber sensing, and particularly, to an optical time-domain reflectometer and method for eliminating interference fading. The reflectometer comprises a laser device, a frequency-division multiplexing modulator, circulator and signal processing device. The frequency-division multiplexing modulator comprises first and second modulators, first and second attenuators. The laser device is configured to generate a coherent light source, which source sequentially passes through the first modulator and the first attenuator to obtain a first optical pulse signal, sequentially passes through the first and second modulators to obtain a second optical pulse signal, and sequentially passes through the second attenuator and the second modulator to obtain a third optical pulse signal.

Abstract: A method for processing a substrate loading the substrate onto a plurality of lift pins passing through a pedestal arranged in a processing chamber. The plurality of lift pins are lowered to rest the substrate on the pedestal. A deposition gas mixture is supplied to deposit film on the substrate. The supply of the deposition gas mixture is stopped. The substrate is raised above the pedestal in the processing chamber using the plurality of lift pins. An etch gas mixture is supplied. Plasma is struck in the processing chamber between the substrate and the surface of the pedestal to remove residual film.

Abstract: Embodiments of the disclosure provide a method for index determination, a decoder, an encoder, and a bitstream. The method includes the following. A first index of a current node is determined based on occupied child nodes of at least one decoded neighbouring node of the current node on a plane perpendicular to a k-th axis.

Abstract: A halidosilyl-substituted cyclic silicon precursor compound has at least 2 to 1 of carbon to silicon ratio and is defined by Formula I herein. A method for forming a film comprising silicon and oxygen and having a carbon content ranging from 10 at. % to 50 at. % via a thermal ALD process includes placing one or more substrates comprising a surface feature into a reactor; heating to reactor to one or more temperatures ranging from ambient temperature to about 600° C.

Abstract: A method includes following steps. A pressure within a chamber is reduced from a first pressure level to a second pressure level using a first pump. After reducing the pressure within the chamber to the second pressure level, a plasma etching operation is performed in the chamber. After performing the plasma etching operation, a pneumatic purging apparatus is placed onto the first pump. After placing the pneumatic purging apparatus onto the first pump, an interior space of the first pump is purged by ejecting a purging gas from the pneumatic purging apparatus.

Abstract: In a method for monitoring a plasma output from a plasma device used to clean an object, a first wavelength of light emitted by the plasma output to be monitored is identified and an intensity of the first wavelength of light is sensed. A second wavelength of light emitted by the plasma output is also identified to be used as a baseline, and an intensity of the second wavelength of light is sensed. The intensities of the first and second wavelengths of light are then compared to determine whether their relationship falls within an acceptable threshold.

Abstract: Systems and methods for providing assistance information related to a sidelink connection are provided. A wireless device is configured to determine a condition associated with a sidelink connection between the wireless device and a second wireless device. The condition can be related to an instability of the sidelink connection. Assistance information is transmitted to a network node in accordance with determining the condition associated with the sidelink connection. One or more sidelink positioning measurements can be modified accordingly.

Abstract: A substrate processing system includes a process chamber, processing stations, and a hub. The processing stations are disposed in the process chamber. Each of the processing stations includes a respective substrate support. The hub is disposed between the processing stations. The hub rotates between a first position and a second position, directs a gas to clean a first area of the process chamber when in the first position, and directs the gas to clean a second area of the process chamber when in the second position.

Abstract: The present disclosure relates to a method performed by a User Equipment (UE) for adapting procedures in a communication system. The UE is an Air-to-ground (ATG) capable UE. The UE is configured for performing one or more measurements on one or more cells of one or more carrier frequencies. The UE is served by a network node in an ATG cell. The UEU determines whether or not the UE is located at least substantially at a cell center of the ATG cell by checking if one or more criteria are met or not. The UE adapts one or more procedures based on a result of the determining.

Abstract: Disclosed are a method, a device, a system and a medium for controlling the ornament lighting, and the method includes the following steps: connecting an ornament to obtain identification codes of the ornament; retrieving stored information about ornament scenes and light-emitting bodies according to the identification codes of the ornament; displaying the ornament scenes visually; selecting the light-emitting bodies of the ornament scenes for setting, and sending the setting information and the codes corresponding to the light-emitting bodies to the ornament; and having the ornament adjusted states of the corresponding light-emitting bodies according to the received setting information and codes.

Abstract: A method of small data transmission (SDT) and related wireless device are described. The method comprises obtaining information at a first time about one or more pre-configured SDT resources including a timing advance (TA) value. A validity of a first measurement of the serving cell is determined by determining whether it was obtained within a first time range based on the first time and on a first set of one or more 5G New Radio (NR) parameters. A validity of a second measurement of the serving cell is determined by determining whether it was obtained within a second time range based on a second time (start time of next SDT occasion), and on a second set of one or more 5G NR parameter. The method further comprises validating the TA when the first and second measurements are valid, and performing a SDT uplink transmission using the validated TA.