Abstract: An information search method includes: obtaining search words at least including a question to be searched and obtaining an initial text vector representation of the search words; obtaining a video corresponding to the search words, and obtaining multi-modality vector representations of the video; starting from the initial text vector representation, performing N rounds of interaction between the video and the search words based on the multi-modality vector representations and a text vector representation of the search words of a current round, to generate a target fusion vector representation, where N is an integer greater than or equal to 1; and obtaining target video frames matching the question to be searched by annotating the video based on the target fusion vector representation.

Abstract: The present application provides a light sensing method and a light sensor module thereof, wherein the light sensing method includes obtaining optical signal values respectively by sensing of a first photodetection element, a second photodetection element, and a third photodetection element, wherein the first photodetection element and the second photodetection element have different effective photosensitive wavelength ranges, and the third photodetection element is a full-absorption band photodetection element. Through this, operation results of the optical signal values may be used to identify types of light sources.

Abstract: A vehicle lamp device includes an outer frame defining an accommodation space, a light shaping frame, a light-emitting unit, and a light distributing cover. The light shaping frame divides the accommodation space into scattering zones, having first and second light shaping regions spaced apart from each other. The light-emitting unit includes light-emitters aligned with the scattering zones. The light-emitters have first and second light-emitting modules for emitting light towards the first and second light shaping regions. The light distributing cover has lens portions aligned with the scattering zones, and allow illumination of light. Each lens portion has a first lens section that is aligned with the light shaping region of the respective scattering zone, and a second lens section that is aligned with the second light shaping region of the respective scattering zone. The first lens section has a lens curvature different from that of the second lens section.

Abstract: A text translation method is described that includes initially acquiring text. Thereafter, first text is determined in the initial text; and second text is determined according to the first text, where the second text is used for describing the first text. Additionally, initial text is translated to obtain initial translation text, and the second text is translated to obtain description translation text. Thereafter, the initial translation text is updated according to the description translation text to obtain target translation text of the initial text.

Abstract: A method for detecting slow HTTP DoS (SHD) attacks in a backbone network can detect three different types of SHID attacks. The method is divided into an off-line training phase and an on-line detection phase. In the off-line training phase, several types of representative unidirectional traffic features are extracted according to attack characteristics of different SHD types and corresponding feature groups are built, where these features can effectively deal with a large amount of unidirectional traffic in backbone networks; a public backbone network dataset is systematically sampled and data are stored in combination with Count-min Sketch, which greatly minimizes storage and computational overhead required in the backbone networks; and a specific machine learning algorithm is used for training to obtain attack detection models. The method can be used for detecting and warning SHD attacks in mass traffic scenarios such as backbone networks to provide a basis for maintaining network security.

Abstract: A method for pre-training a language model includes: constructing a pre-training language data set, in which the pre-training language data set comprises unsupervised language data and supervised language data; generating a hierarchical multi-template and multi-task language data set based on the pre-training language data set; and pre-training the language model based on the hierarchical multi-template and multi-task language data set.

Abstract: A power generating wheel rim device includes a wheel rim body, a power generating mechanism, a cover mechanism, and a circuit mechanism. The wheel rim body includes a central plate and a central slot recessed into the central plate. The power generating mechanism is embedded in the central slot and is operable to be driven by the wheel rim body to rotate and to generate electricity. The cover mechanism covers the power generating mechanism, is mounted to the central plate of the wheel rim body, and includes an air permeable passage structure that is annular and disposed along a periphery of the cover mechanism and that is in 10 spatial communication with an external environment and the central slot. The circuit mechanism is mounted in the cover mechanism and is electrically connected to the power generating mechanism for converting alternating current of the electricity generated into direct current.

Abstract: Methods of forming magnetic tunnel junction (MTJ) memory cells used in a magneto-resistive random access memory (MRAM) array are provided. A pre-clean process is performed to remove a metal oxide layer that may form on the top surface of the bottom electrodes of MTJ memory cells during the time the bottom electrode can be exposed to air prior to depositing MTJ layers. The pre-clean processes may include a remote plasma process wherein the metal oxide reacts with hydrogen radicals generated in the remote plasma.

Abstract: A power generating device is adapted to be mounted to a rotatable object, and includes a housing adapted to be fixedly mounted to the object, a rotor module fixedly and coaxially mounted to the housing along an axial line that is substantially horizontal, a stator module mounted to the housing and rotatable relative to the housing about the axial line, and a counterweight module fixedly mounted to the stator module. A center of gravity of the counterweight module is located below the axial line. When the object rotates, the rotor module is driven to rotate relative to the stator module which does not rotate with the rotor module due to a force of gravity of the counterweight module, and one of the rotor module and the stator module generates electric power according a magnetic field provided by another one of the rotor module and the stator module through electromagnetic induction.

Abstract: A signal sensing device includes a body and two signal sensing elements disposed in the body. An insulating layer is sandwiched between the two signal sensing elements. Each of the two signal sensing elements incudes a signal transmission section and a signal sensing section in electrical connection with the signal transmission section. The signal transmission sections are planar antennae parallel to each other and each having an antenna shape of meander-line type. The antenna shape of each transmission section has a vertical projection on a plane parallel to each signal transmission section. The vertical projections of the antenna shapes do not overlap completely. When a portion of the body forms a surrounding portion which surrounds a to-be-sensed target, a portion or an entirety of each signal sensing section is located on the surrounding portion.

Abstract: Described are methods of measuring a concentration of a photosensitive group in a liquid, for example a concentration of amine-based inhibitor in an etching solution, and related equipment and systems.