Abstract: An information push method, includes acquiring a photo of an environment in which a client is located; recognizing a target object from the photo and collecting statistics on a frequency value or a frequency of appearance; and pushing related information of the target object to the client according to the frequency value or the frequency. An information push system, a digital television receiving terminal, and a computer storage medium are further provided. Statistics are collected on quantized data of how often a target object appears in an environment in which a client is located by using a scientific and technical means and the statistical data is used as a reference for push.

Abstract: A disk array flushing method and a disk array flushing apparatus. The method includes acquiring a sequence, which is according to physical addresses of logical units, of the logical units in a same redundant array of independent disks (RAID) group in a disk array, and sequentially flushing, according to the sequence of the logical units, concurrent flushing input-outputs (IOs) of the RAID group to the logical units in the RAID group. According to the disk array flushing method and the disk array flushing apparatus, by uniformly scheduling logical units in a single RAID group and performing flushing according to a sequence of physical addresses, it takes less time for a magnetic arm to jump back and forth for addressing, and influences between RAID groups are prevented by means of independent control over the RAID groups, thereby improving flushing efficiency of a disk array, and increasing throughput of the disk array.

Abstract: Disclosed herein is an X-ray imaging system, comprising: a first X-ray source; an image sensor, spaced apart from the first X-ray source; a table configured to be positioned between the first X-ray source and the image sensor and configured to accommodate a person for imaging in the X-ray imaging system; and a fiducial marker system; wherein the fiducial marker system comprises three markers, centers of the three markers not being collinear; wherein at least one of the three markers is distinct. The X-ray imaging system may be used to model the 3D shape and track the 3D motion of an object (e.g., a surgical instrument) in the person.

Abstract: A retractable power plug includes a housing having an opening formed therein, a mounting portion fixed in the housing, a plurality of pins, and a cover plate. The pins are rotatably mounted on the mounting portion. The cover plate at least partially cover the opening and defines a plurality of receiving grooves corresponding to the pins. The plurality of pins are configured in a first configuration and protrude out of the receiving grooves and in a second configuration to be within the plurality of receiving grooves. The cover plate further defines at least one recess on a sidewall of one of the plurality of receiving grooves, thereby enabling the pin to be retracted.

Abstract: Systems and methods are provided for speech recognition. For example, audio characteristics are extracted from acquired voice signals; a syllable confusion network is identified based on at least information associated with the audio characteristics; a word lattice is generated based on at least information associated with the syllable confusion network and a predetermined phonetic dictionary; and an optimal character sequence is calculated in the word lattice as a speech recognition result.

Abstract: An active plasmon sensor comprising a single crystalline semiconductor CdS nano-slab atop a silver surface separated by a magnesium fluoride (MgF2) gap layer. The surface plasmon effect localizes the electromagnetic field at the interface between the metal and semiconductor, allowing both the device's physical size and mode confinement to shrink down to the nanometer scale in a dimension perpendicular to the metal surface.

Abstract: The present invention provides a polysilicon manufacturing method that enhances homogeneity of a polysilicon layer, including (1) forming a amorphous silicon layer (30) on a substrate (10); (2) dividing the amorphous silicon layer (30) into a plurality of areas and measuring a film thickness of the amorphous silicon layer (30) in each of the areas: (3) comparing the measured film thickness of each of the areas with a predetermined film thickness so as to identify and define ones of the areas of which the measured film thicknesses are greater than a predetermined film thickness as excessive-film-thickness areas (42); (4) spraying an etchant liquid to rinse the amorphous silicon layer (30) of the excessive-film-thickness areas (42) in order to etch off a portion of the amorphous silicon layer (30) of each of the excessive-film-thickness areas (42) and at the same time, spraying pure water to rinse remaining ones of the areas so as to make the film thickness of each of the areas of the amorphous silicon layer (3

Abstract: A method is performed at a device having one or more processors and memory. The device establishes a first-level Deep Neural Network (DNN) model based on unlabeled speech data, the unlabeled speech data containing no speaker labels and the first-level DNN model specifying a plurality of basic voiceprint features for the unlabeled speech data. The device establishes a second-level DNN model by tuning the first-level DNN model based on labeled speech data, the labeled speech data containing speech samples with respective speaker labels, wherein the second-level DNN model specifies a plurality of high-level voiceprint features. Using the second-level DNN model, registers a first high-level voiceprint feature sequence for a user based on a registration speech sample received from the user. The device performs speaker verification for the user based on the first high-level voiceprint feature sequence registered for the user.

Abstract: A method and a device for training a DNN model includes: at a device including one or more processors and memory: establishing an initial DNN model; dividing a training data corpus into a plurality of disjoint data subsets; for each of the plurality of disjoint data subsets, providing the data subset to a respective training processing unit of a plurality of training processing units operating in parallel, wherein the respective training processing unit applies a Stochastic Gradient Descent (SGD) process to update the initial DNN model to generate a respective DNN sub-model based on the data subset; and merging the respective DNN sub-models generated by the plurality of training processing units to obtain an intermediate DNN model, wherein the intermediate DNN model is established as either the initial DNN model for a next training iteration or a final DNN model in accordance with a preset convergence condition.

Abstract: A self-cleaning milk foaming device includes a mixing unit, a steam jet unit and a control unit. The mixing unit has a mixing chamber, and a milk inlet passage, a cleaner inlet passage and a foaming gas inlet passage communicated with the mixing chamber. The mixing chamber has a milk outlet nozzle, an upstream section, and a downstream section disposed between the upstream section and the milk outlet nozzle. The milk inlet passage is situated between the cleaner inlet passage and the milk outlet nozzle. The foaming gas inlet passage is disposed upstream of the milk inlet passage. The steam jet unit is for supplying steam into the mixing chamber. The control unit is for controlling the cleaner to flow or to stop flowing into the cleaner inlet passage.

Abstract: A method and device for voiceprint recognition, include: establishing a first-level Deep Neural Network (DNN) model based on unlabeled speech data, the unlabeled speech data containing no speaker labels and the first-level DNN model specifying a plurality of basic voiceprint features for the unlabeled speech data; obtaining a plurality of high-level voiceprint features by tuning the first-level DNN model based on labeled speech data, the labeled speech data containing speech samples with respective speaker labels, and the tuning producing a second-level DNN model specifying the plurality of high-level voiceprint features; based on the second-level DNN model, registering a respective high-level voiceprint feature sequence for a user based on a registration speech sample received from the user; and performing speaker verification for the user based on the respective high-level voiceprint feature sequence registered for the user.

Abstract: An apparatus and a method for detecting a buffer overflow attack, and a security protection system. The apparatus for detecting a buffer overflow attack includes a memory storing instructions, a processor configured to execute the instructions stored in the memory to obtain external input data for a target process, determine that the target process decodes the external input data, detect attack code on the decoded external input data, wherein the attack code is a code used for performing an overflow attack on a buffer, where the apparatus or the method facilitates detection of attack code from the data obtained by decoding, and may improve a detection rate of the attack code.

Abstract: A method of recognizing speech is provided that includes generating a decoding network that includes a primary sub-network and a classification sub-network. The primary sub-network includes a classification node corresponding to the classification sub-network. The classification sub-network corresponds to a group of uncommon words. A speech input is received and decoded by instantiating a token in the primary sub-network and passing the token through the primary network. When the token reaches the classification node, the method includes transferring the token to the classification sub-network and passing the token through the classification sub-network. When the token reaches an accept node of the classification sub-network, the method includes returning a result of the token passing through the classification sub-network to the primary sub-network. The result includes one or more words in the group of uncommon words. A string corresponding to the speech input is output that includes the one or more words.

Abstract: An electronic device with one or more processors and memory trains an acoustic model with an international phonetic alphabet (IPA) phoneme mapping collection and audio samples in different languages, where the acoustic model includes: a foreground model; and a background model. The device generates a phone decoder based on the trained acoustic model. The device collects keyword audio samples, decodes the keyword audio samples with the phone decoder to generate phoneme sequence candidates, and selects a keyword phoneme sequence from the phoneme sequence candidates. After obtaining the keyword phoneme sequence, the device detects one or more keywords in an input audio signal with the trained acoustic model, including: matching phonemic keyword portions of the input audio signal with phonemes in the keyword phoneme sequence with the foreground model; and filtering out phonemic non-keyword portions of the input audio signal with the background model.

Abstract: In a method of manufacturing a spindle motor, a sheet preferably made of a thermoplastic resin is closely adhered to an upper surface of a base portion of a spindle motor such that the sheet covers a base hole of the base portion. Next, heat is applied to a region of the sheet which coincides with the base hole to define a sheet hole continuous with the base hole. Thereafter, a coil is arranged over the base portion, and a conducting wire defining the coil is extended through the sheet hole and the base hole to be drawn downwardly out of the base portion. The sheet hole is thus defined after the sheet is closely adhered to the base portion. In addition, the sheet hole is defined by melting the resin through application of the heat. It is therefore easy to arrange the sheet and define the sheet hole.

Abstract: The present disclosure discloses a method of sharing information with digital television (“DTV”) terminals and terminal and system for the same, the method comprising: establishing a communication connection between a DTV terminal and a mobile terminal; sending from the DTV terminal to the mobile terminal the DTV terminal's data receiving port information; and sending one or more multimedia files from the mobile terminal to the DTV terminal in accordance with the DTV terminal's data receiving port information. The present disclosure can accomplish information sharing between mobile terminals and DTV terminals, thereby realizing easy and convenient browsing or playback of multimedia files, taking advantage of DTVs' superior multimedia playback effect and avoiding the inconvenience brought about by wired connections.

Abstract: This disclosure provides systems, methods, and apparatus related to nanophotonics. In one aspect, an arrangement of waveguides includes a substrate and three waveguides. Each of the three waveguides may be a linear waveguide. A second waveguide is positioned between a first waveguide and a third waveguide. The dimensions and positions of the first, the second, and the third waveguides are specified to substantially eliminate coupling between the first waveguide and the third waveguide over a distance of about 1 millimeter to 2 millimeters along lengths of the first waveguide, the second waveguide, and the third waveguide.

Abstract: A method and system for adding punctuation to a voice file is disclosed. The method includes: utilizing silence or pause duration detection to divide a voice file into a plurality of speech segments for processing, the voice file includes a plurality of features units; identifying all features units that appear in the voice file according to every term or expression and semantics features of the every term or expression that form each of the plurality of speech segments; using a linguistic model to determine a sum of weight of various punctuation modes in the voice file according to all the feature units, the linguistic model is built upon semantics features of various parsed out terms or expressions from a body text of a spoken sentence according to a language library; and adding punctuations to the voice file based on the determined sum of weight of the various punctuation modes.

Abstract: A method and a device for training an acoustic language model, include: conducting word segmentation for training samples in a training corpus using an initial language model containing no word class labels, to obtain initial word segmentation data containing no word class labels; performing word class replacement for the initial word segmentation data containing no word class labels, to obtain first word segmentation data containing word class labels; using the first word segmentation data containing word class labels to train a first language model containing word class labels; using the first language model containing word class labels to conduct word segmentation for the training samples in the training corpus, to obtain second word segmentation data containing word class labels; and in accordance with the second word segmentation data meeting one or more predetermined criteria, using the second word segmentation data containing word class labels to train the acoustic language model.

Abstract: A method includes: acquiring data samples; performing categorized sentence mining in the acquired data samples to obtain categorized training samples for multiple categories; building a text classifier based on the categorized training samples; classifying the data samples using the text classifier to obtain a class vocabulary and a corpus for each category; mining the corpus for each category according to the class vocabulary for the category to obtain a respective set of high-frequency language templates; training on the templates for each category to obtain a template-based language model for the category; training on the corpus for each category to obtain a class-based language model for the category; training on the class vocabulary for each category to obtain a lexicon-based language model for the category; building a speech decoder according to an acoustic model, the class-based language model and the lexicon-based language model for any given field, and the data samples.