Abstract: The disclosure relates to a video anomaly detection method based on human-machine cooperation, in which video frames and traditional descriptors of optical stream of an image are utilized as an input for auto-encoder neural network coding, and converted into a representation content of a hidden layer, and then the representation content of the hidden layer is decoded, reconstructed and output. The auto-encoder network is trained with normal samples. In a test stage, if an input is a normal sample, a final reconstructed error keeps high similarity with an input sample; on the contrary, if the input is an abnormal sample, the final reconstructed error deviates greatly from the input sample.

Abstract: A method for extracting spectral information of a substance under test includes: identifying a pixel region A(x, y) occupied by an object under test from a hyperspectral image acquired; extracting a specular reflection region Aq and a diffuse reflection region Ar from the pixel region A(x, y), and calculating a representative spectrum Iq(?) of the specular reflection region Aq and a representative spectrum Ir(?) of the diffuse reflection region Ar, respectively; by comparing each element in the representative spectrum Iq(?) of the specular reflection region Aq with each element in the representative spectrum Ir(?) of the diffuse reflection region Ar, separating information of a light source from spectral information of the object to obtain a first spectral invariant C(?). This method does not require additional spectral information of the light source, which improves the analysis efficiency.

Abstract: Disclosed are method and apparatus for predicting hard disk fault occurrence time of hard disk failure, and storage medium. The method includes steps of: screening a hard disk on the verge of failure from a plurality of hard disks according to state data acquired of hard disks; calculating variation quantity and discrete quantity of each piece of the state data of the hard disk on the verge of failure acquired over a first preset period of time, to obtain a first predicted data set; and inputting the first predicted data set into a first training model to obtain probability of occurrence of failure for the hard disk on the verge of failure over a future second preset period of time.

Abstract: The embodiments of the present application provides a target detection method, which comprises: obtaining a plurality of images acquired by an image acquiring device as images to be detected when a region to be detected is illuminated by light sources with different illuminating angles, wherein the different illuminating angles correspond to the different images to be detected (S101); and determining whether a target exists in the region to be detected based on gray scale differences between the obtained images to be detected with illuminating by the light sources of with different illuminating angles (S102). The method is applied to achieve a simple and effective detection of the target existing in the region to be detected.

Abstract: A system for testing a spectral response speed of a tunable filter is disclosed, which includes a collimating light source, a beam splitting element, a focusing lens, and an image recording device of light spot position arranged successively. The tunable filter is disposed between the collimating light source and the beam splitting element and configured to be continuously tuned within a certain wavelength range during testing. The beam splitting element is used to form light beams of different wavelength bands passing through the tunable filter into diffracted beams or refracted beams corresponding to different wavelength bands. The focusing lens is used to perform focusing. The image recording device of light spot position is used to record change information about positions where the diffracted beams or refracted beams corresponding to different wavelength bands are imaged.

Abstract: A data distribution system includes: an edge stream-pushing server configured to receive streaming media data packets sent by a stream pusher, serialize the streaming media data packets, and forward the serialized streaming media data packets to a plurality of central servers; a central server configured to cache the serialized streaming media data packets, if a serialized back-to-source request from the edge stream-pulling server is received, determine, a start position of to-be-transmitted streaming media data packets according to the serialized back-to-source request, and send the cached streaming media data packets to the edge stream-pulling server according to the start position; and an edge stream-pulling server configured to deserialize received streaming media data packets to restore the received streaming media data packets, and send the restored streaming media data packets to a user terminal.

Abstract: A wear leveling and access method and device for a non-volatile memory, and a storage medium. The method includes: logically dividing a non-volatile memory into physical units of p levels, the non-volatile memory including a plurality of physical units of the first level, each physical unit of the p?1-th level including a plurality of physical units of the p-th level, and p being a positive integer greater than one (S110); when a time period corresponding to the physical units of the q-th level arrives, replacing the data of each of the physical units of the q-th level with other physical units of the q-th level, q being any positive integer from one to p (S120).

Abstract: A filter is irradiated with incident light, and calibrated light with a fixed wavelength is obtained after the incident light is transmitted through a calibration filter film on the filter; the calibrated light is selected and transmitted via the Fabry-Perot cavity, and selected light is obtained after the calibrated light is transmitted through the Fabry-Perot cavity; an image chip is irradiated with the selected light for imaging, and an output signal is obtained. When the selected wavelength of the tuneable Fabry-Perot cavity matches the wavelength of the calibrated light transmitted by the calibration filter film at a certain position, light at the position is selected and reaches the imaging chip, and the output signal of the imaging chip at the corresponding position can be used as a detection signal for self-calibration, which does not affect normal imaging at other positions on the filter besides the calibration filter film.

Abstract: A semiconductor device includes a patterned metal layer on a substrate, via conductors on the patterned metal layer, first inter-metal dielectric (IMD) patterns embedded in the patterned metal layer, and a second IMD pattern surrounding the patterned metal layer. Preferably, the first IMD patterns are between and without overlapping the via conductors in a top view.

Abstract: Laser cleaning equipment and a cleaning method for a shaft component are provided. The equipment includes: a supporting base assembly; two driving wheel structures on the supporting base assembly, driving wheels of each of which is configured for being close to or away from each other, and the shaft component to be cleaned is placed between the two driving wheel structures; a friction wheel structure that is tangent to driving wheel structure(s) and uses a friction force thereof to drive driving wheel structure(s) to rotate; a connection shaft assembly that coaxially passes through the friction wheel structure; a power driving mechanism, one end of which that faces toward the connection shaft assembly is in drive connection with the connection shaft assembly and is configured to drive the connection shaft assembly to rotate; and a laser cleaning mechanism configured for performing laser cleaning on the shaft component to be cleaned.

Abstract: A method and an apparatus for storing, reading, and displaying a plurality of multimedia files are disclosed.

Abstract: A wafer-level integrated thermal detector comprises a first wafer and a second wafer (W1, W2) bonded together. The first wafer (W1) includes a dielectric or semiconducting substrate (100), a dielectric sacrificial layer (102) deposited on the substrate, a support layer (104) deposited on the sacrificial layer or the substrate, a suspended active element (108) provided within an opening (106) in the support layer, a first vacuum-sealed cavity (110) and a second vacuum-sealed cavity (106) on opposite sides of the suspended active element. The first vacuum-sealed cavity (110) extends into the sacrificial layer (102) at the location of the suspended active element (108). The second vacuum-sealed cavity (106) comprises the opening of the support layer (104) closed by the bonded second wafer.

Abstract: A method for fabricating semiconductor device includes the steps of first forming a first inter-metal dielectric (IMD) layer on a substrate, patterning the first IMD layer to form first IMD patterns on the substrate, a trench surrounding the first IMD patterns, and a second IMD pattern surrounding the trench, forming a metal layer in the trench to surround the first IMD patterns, forming a second IMD layer on the first IMD patterns, the metal layer, and the second IMD pattern, and forming via conductors in the second IMD layer. Preferably, the via conductors not overlapping the first IMD patterns.

Abstract: A data distribution system includes: an edge stream-pushing server configured to receive streaming media data packets sent by a stream pusher, serialize the streaming media data packets, and forward the serialized streaming media data packets to a plurality of central servers; a central server configured to cache the serialized streaming media data packets, if a serialized back-to-source request from the edge stream-pulling server is received, determine, a start position of to-be-transmitted streaming media data packets according to the serialized back-to-source request, and send the cached streaming media data packets to the edge stream-pulling server according to the start position; and an edge stream-pulling server configured to deserialize received streaming media data packets to restore the received streaming media data packets, and send the restored streaming media data packets to a user terminal.

Abstract: Provided are a file sending method, a file receiving method, and a file transceiving apparatus. The file sending method includes: creating a data sending process; acquiring a file to be sent, corresponding to the data sending process, from a data cache region in a user space; constructing metadata for the data sending process, and recording the metadata in a data reading and sending region in the user space; and sending the file to be sent and the metadata to a network adapter by means of a user-mode network device driver, and sending the file to be sent and the metadata by means of the network adapter.

Abstract: Methods, devices, and systems for identifying charging devices are provided. In one aspect, a method of identifying a charging device include: capturing an infrared image and a depth image of a current field of view with a depth camera; determining, according to the infrared image, that there are one or more suspected charging device areas that satisfy first specified conditions; determining, according to the depth image, that there is a target charging device area whose height relative to a depth camera is within a specified range in the one or more suspected charging device areas; and identifying the charging device according to the target charging device area. The first specified conditions indicate that a gray-scale value of each of pixels in an area is greater than a second specified value, and a number of the pixels in the area is greater than a third specified value.

Abstract: A wear leveling and access method and device for a non-volatile memory, and a storage medium. The method includes: logically dividing a non-volatile memory into physical units of p levels, the non-volatile memory including a plurality of physical units of the first level, each physical unit of the p?1-th level including a plurality of physical units of the p-th level, and p being a positive integer greater than one (S110); when a time period corresponding to the physical units of the q-th level arrives, replacing the data of each of the physical units of the q-th level with other physical units of the q-th level, q being any positive integer from one to p (S120).

Abstract: A method for fabricating semiconductor device includes the steps of first forming a first inter-metal dielectric (IMD) layer on a substrate, patterning the first IMD layer to form first IMD patterns on the substrate, a trench surrounding the first IMD patterns, and a second IMD pattern surrounding the trench, forming a metal layer in the trench to surround the first IMD patterns, forming a second IMD layer on the first IMD patterns, the metal layer, and the second IMD pattern, and forming via conductors in the second IMD layer. Preferably, the via conductors not overlapping the first IMD patterns.

Abstract: Embodiments of the present application provide a method and apparatus for updating an online self-learning event detection model.