117 results about "End point detection" patented technology

The embodiment of the invention discloses an audio processing method and device, and the method comprises the steps: carrying out the traversal of a segmented sentence list which is determined through the interception of an audio stream based on a mute detection result, and determining a current segmented sentence and an endpoint detection result of the current segmented sentence, determining a start frame of the current segmented sentence as a start frame of a latest to-be-identified sentence when the current segmented sentence has a sentence start point, determining an end frame of the current segmented sentence as an end frame of the latest to-be-identified sentence when the current segmented sentence has a sentence end point, and determining the latest to-be-identified sentence according to the start frame and the end frame of the latest to-be-identified sentence. According to the method and the device, the segmented sentences in the audio stream can be continuously updated, the to-be-recognized sentences can be determined according to the endpoint detection results of the segmented sentences, continuous long speech recognition is realized, and the complete content of the to-be-recognized sentences from the start frame to the end frame is recognized, so that the speech recognition efficiency is improved. The accuracy of long speech recognition can be improved.

The invention relates to the technical field of voice recognition intelligent home control, and discloses a non-specific person voice recognition intelligent switch control method and system based on deep learning. The non-specific person voice recognition intelligent switch control method is applied to intelligent switch control equipment, and specifically comprises the following steps of S101, receiving original audio signals continuously collected in a to-be-controlled environment, and preprocessing the collected original audio signals, and then a starting point and an ending point of an effective voice segment are positioned by adopting endpoint detection based on a double-threshold method and combining the characteristic parameters of the short-time energy and the short-time zero-crossing rate. A multi-layer hidden layer structure with Dropout regularization is adopted in a neural network model, the generalization ability of the model is enhanced, a context sensing mechanism is introduced into a semantic understanding module, a composite instruction containing azimuth information can be intelligently analyzed, crossing from recognition to understanding is achieved, and the method has the advantages of being high in practicability and easy to popularize. The system is ensured to maintain a high recognition rate for voice instructions of different users under different environment conditions.

Methods and systems for process chamber qualification for maintenance process endpoint detection are provided. Data collected by one or more sensors of a process chamber of a manufacturing system is identified. The identified data is collected during performance of initial maintenance operation(s) of a maintenance process. A current state of the process chamber is determined, based on the identified data, after the performance of the initial maintenance operation(s) based on the identified data. In response to a determination that the current state does not satisfy one or more chamber maintenance criteria, a set of subsequent maintenance operations to be performed to cause the current state of the process chamber to satisfy the criteria is identified. Performance of the set of subsequent maintenance operations is initiated at the process chamber.

Information processing device that includes: an information acquisition unit that acquires reliability degradation factor state information in a chemical mechanical polishing process of a substrate performed by a substrate processing device, the reliability degradation factor state information including at least one of wear state information indicating a wear state of components of the substrate processing device and processing state information indicating a processing state during polishing; and a state prediction unit that predicts reliability information of a polishing endpoint detection function for the reliability degradation factor state information by inputting the reliability degradation factor state information acquired by the information acquisition unit into a learning model that has been trained through machine learning to learn a correlation between the reliability degradation factor state information and reliability information of the polishing endpoint detection function that indicates reliability of an endpoint detection function that detects that the chemical mechanical polishing process has reached an endpoint.

The present application provides a method, apparatus, device and storage medium for voice endpoint detection, wherein the voice endpoint detection method can determine whether the audio frame contained in the audio data to be detected is a silence frame, a noise frame or a voice frame, that is, the present application can detect the relatively accurate attributes of the audio frame contained in the audio data, and perform detection of the voice front-end point and the voice back-end point on this basis, so as to obtain relatively accurate detection results. On the basis of realizing voice endpoint detection, the present application can obtain the recognition text of the voice segment, and can determine the semantic scene of the recognition text based on the semantics of the recognition text, and then can set a suitable post-silence timeout threshold based on the semantic scene of the recognition text, thereby triggering a post-silence timeout event based on the suitable post-silence timeout threshold to improve the user experience.

The invention relates to the technical field of medical information processing, in particular to a medical voice instruction semantic analysis and form automatic filling system, which comprises the following modules: a voice recognition and analysis module, which is used for collecting a medical voice instruction, carrying out noise reduction processing and endpoint detection on a voice signal, and sending the voice signal to an automatic form filling module; generating a recognition text with confidence by using a voice recognition model customized by a medical scene, and performing named entity recognition and slot extraction on the recognition text to generate candidate semantic units; and the constraint generation module is used for extracting clinical logic constraints of the form fields according to the candidate semantic units. According to the method, constraint semantic representation is generated in combination with violation degree calculation, matching is completed under logic constraints by using a global optimization algorithm, and high-precision mapping of medical voice instructions and form fields is realized; not only is the integrity of semantic information ensured in the recognition stage, but also mistaken filling is effectively avoided through clinical logic constraints in the matching stage.

The application relates to a container endpoint detection method and device, a storage medium and equipment. A least square method is used to construct a straight line equation of n sampling points of laser scanning data of a to-be-detected container. Whether a sampling point after the nth sampling point belongs to the same straight line equation is judged to obtain an offset sampling point. The coordinate information of the first sampling point of the straight line equation and the coordinate information of the offset sampling point are used to obtain the coordinate information of a straight line endpoint. The straight line equation of the sampling point after the offset sampling point is reconstructed, and the coordinate information of the endpoint of the straight line segment corresponding to the straight line equation is obtained. Therefore, the position information of each endpoint of the to-be-detected container can be obtained according to the coordinate information of the endpoint, the user can quickly locate the endpoint position of the container according to the position information of each endpoint, and the detection efficiency of the container is improved.

The semantic perception real-time voice endpoint detection method is applied to a voice interaction system, and comprises the following steps: receiving a voice stream of a user and a voice stream sent by the system from at least two audio channels; performing causal voice coding on the voice stream of the user and the voice stream sent by the system to obtain corresponding double-path voice representation; a causal two-way modeling architecture is constructed, and the causal two-way modeling architecture is realized by adopting an autoregressive prediction network and is used for receiving two-way voice representation and modeling a speech round conversion dynamic state between a user and a system through an embedded causal constraint mechanism and a cross-channel interaction mechanism; based on the output of the autoregressive prediction network, jointly predicting the category of a speech round conversion event and the remaining time from the current moment to the end of the speech round; and a smoothing processing mode is used to correct the predicted speech wheel conversion event type, and a final endpoint detection result is determined. According to the method, the endpoint detection stability in a complex dialogue scene can be improved.

This invention discloses a method for accurate audio warning identification based on hybrid features. This method analyzes collected audio speech to determine the issuance time of the warning signal, thereby accurately evaluating the timeliness of the audio warning. First, a double noise reduction method using logmmse-spectral subtraction is employed to filter out noise information in the recorded speech. Next, endpoint detection based on short-time energy is used to mark the effective speech segments in the test speech. Then, MFCC features and waveform polynomial features are extracted from each frame of the effective speech segment. Subsequently, the two features are used as inputs to two channels of a convolutional neural network, and the outputs of the two channels are summed to obtain the hybrid features. Finally, the hybrid features are used as input to a softmax function, and the speech segment containing the target speech (audio warning signal) is determined by the maximum probability value. The starting position of this speech segment is the issuance time of the warning signal.

To provide a method for manufacturing a polishing pad with an endpoint detection window (window member) that exhibits excellent adhesion between the back surface of the window member and the adhesive surface of the second sheet. [Solution] A polishing pad comprising a first sheet having a first through-hole, a second sheet having a second through-hole, and a window member, wherein the first sheet has a polishing surface for polishing an object to be polished, the second sheet is bonded to the surface of the first sheet opposite to the polishing surface, the outer periphery of the second through-hole is formed to cover the outer periphery of the first through-hole, and the window member is positioned inside the first through-hole and on the second sheet such that the uppermost part of the window member is lower than the polishing surface; and a method for manufacturing a polishing pad comprising the steps of: preparing the polishing pad, and placing a spacer of a predetermined thickness on a press platen under the polishing pad around the polishing pad, and pressing the polishing pad with a press machine.

The utility model provides a light acquisition mechanism for an etching machine end point detection system, which relates to the field of etching machines and comprises a supporting seat, a convex lens assembly and an optical filter assembly. The supporting seat comprises a containing cylinder; the convex lens assembly and the optical filter assembly are detachably fixed in the accommodating cylinder; the optical filter assembly comprises a shell with a cylindrical through hole, an optical filter and a clamp spring; an annular baffle is arranged at the opening part of the cylindrical through hole; an annular groove is formed in the inner wall of the cylindrical through hole; the inner diameter of the ring of the snap spring is smaller than the diameter of the cylindrical through hole of the shell; the clamp spring is detachably fixed in the groove; the optical filter is arranged between the clamp spring and the circular ring baffle plate; and the two cylindrical bottom surfaces of the optical filter respectively abut against the snap springs and are fixed with the circular ring baffle plate. The light acquisition mechanism for the etching machine end point detection system provided by the utility model solves the problem that an optical filter is inconvenient to replace in the prior art, and the optical filter can be conveniently and stably replaced.

The application belongs to the field of mechanical processing, and discloses an endpoint detection method and system for polishing of high-hard-brittle materials, which comprises the following steps: step 1, constructing a mathematical model of chemical mechanical polishing; step 2, building a test platform to collect force and torque signals; step 3, designing a circuit of an edge computing platform; step 4, performing a parameter control experiment; step 5, developing a host computer system; step 6, finally obtaining a training set of a high-hard-brittle material processing state prediction model based on force and torque signals; step 7, training the processing state prediction model; and step 8, using the trained processing state prediction model to perform endpoint detection. The application uses polishing torque to perform online real-time detection on the polishing state and uses a state feedback method, so that the polishing processing efficiency can be improved, the polishing quality and the polishing yield can be improved, and the cost can be reduced.

The application provides a regulation voice segmentation method based on speech recognition and endpoint detection, which is applied to air traffic control voice audio stream segmentation, and comprises the following steps: step 1, constructing a punctuation model based on speech recognition and a speech endpoint detection model; step 2, using the punctuation model based on speech recognition to recognize the audio data stream of the regulation voice, and outputting the corresponding text and sentence end identifier of the audio data stream; step 3, using the speech endpoint detection model to judge the speech starting point and ending point contained in the audio data stream of the regulation voice; step 4, segmenting the audio data stream of the regulation voice into audio segments; and step 5, applying the audio segments as data materials to the speech recognition process of the air traffic control system. Through the combination of speech recognition and endpoint detection, the application realizes the automatic segmentation of the air traffic control voice audio stream, and improves the accuracy and efficiency of the segmentation.

An endpoint detection system for enhanced spectral data collection is provided. An optical bundle is coupled to a light source. The optical bundle includes an emitting optical fiber and a receiving optical fiber disposed at a pairing angle relative to the emitting optical fiber. The optical bundle is coupled to a collimator assembly that receives a light beam of incident light from the emitting optical fiber and directs spectral components of the light beam to first and second portions of a substrate surface. The collimator collects reflected spectral components produced by the spectral components directed to the substrate surface. The collimator assembly transmits the reflected spectral components to the receiving fiber, which transmits the reflected spectral components to a light detection component. A processing device coupled to the light detection component determines a reflectance of the substrate surface based on the reflected spectral components.

A technique for threat response associated with an endpoint detection and response (EDR) system. The system uses a combination of automated observable detection, threat intelligence enrichment, graph

The invention discloses a livestock respiratory rate estimation method, and relates to the technical field of animal health monitoring. The method comprises the following steps: collecting a livestock lung sound stream, segmenting into segments according to a fixed window length, and carrying out initial noise reduction; framing endpoints of the segments are detected, starting points and ending points of inspiration segments are marked, and after a sequence is obtained, a small subset and a large subset are divided according to the number of the inspiration segments; screening the effective inspiration section by using the expiration average duration time in the less set, and screening the effective inspiration section by using the difference value between the time interval of two adjacent inspiration sections and the expiration average duration time and the standard difference value in the more set; after the effective inspiration sections are screened out, the breathing frequency is calculated according to the ratio of the time interval between the head and tail inspiration sections of the target section to the number of effective breathing cycles. The accuracy of livestock respiratory rate detection can be effectively improved.

The application provides a speech endpoint detection method, device and equipment and a storage medium. The speech endpoint detection method comprises the following steps: obtaining target data, wherein the target data comprises target audio data; inputting the target data into a pre-trained audio frame classification model to obtain a classification result of an audio frame of the target audio data, wherein the audio frame classification model is trained by taking first audio data with noise in one or more categories of a plurality of noise categories as a training sample, taking a real category of the audio frame of the first audio data in a plurality of dimensions as a sample label, and the plurality of dimensions comprise a speech dimension and noise dimensions corresponding to the plurality of noise categories; and determining a speech endpoint according to the classification result of the audio frame of the target audio data. The speech endpoint detection method provided by the application can detect an accurate speech endpoint.

The purpose of the present application is to provide an endpoint detection method and device for audio, which obtains the target audio to be processed and its corresponding total duration; based on the total duration of the target audio, sequentially performs frequency domain transformation, feature extraction and effective feature filtering on the target audio to obtain the effective feature vector and sampling sequence number corresponding to each sampling moment in the target audio, and the absolute time difference between two adjacent sampling moments; based on the effective feature vector corresponding to each sampling moment, respectively predicts the category of each sampling moment and the relative deviation between each sampling moment and the moment of the true endpoint of the target audio; based on the absolute time difference and the sampling sequence number, category and relative deviation between the moment and the true endpoint of the target audio corresponding to each sampling moment, determines the moment of the starting endpoint and the moment of the ending endpoint of the target audio, thereby realizing accurate endpoint detection of audio with strong noise interference, with low computational complexity and high speed.

A technique for threat response associated with an endpoint detection and response (EDR) system. The system uses a combination of automatic observable detection, threat intelligence augmentation, graph analysis, and supervised machine learning to machine-predict analyst behavior in classifying EDR alerts (as "true" or "false" positives) and supports either (i) automatic suppression of alerts that the system classifies with sufficient confidence as either true or false, or (ii) providing recommendations to analysts for alerts that cannot be so classified to expedite their actions. The automatic detection of observables for graph-based feature detection, coupled with automatic processing of alerts when possible, significantly reduces the overall analyst workload for the EDR system. Furthermore, and even when the machine-based predictions are not sufficiently reliable to allow bypassing the analyst, the system provides additional context and augmentation to the analyst, facilitating faster (or at least more efficient) alert handling.

A method of determining an End Point Detection (EPD) time and an electronic device for performing the method are provided. The electronic device includes a processor and a memory electrically connected to the processor and configured to store commands executed by the processor. The processor may be configured to recognize an utterance when the commands are executed, determine an End Point Detection (EPD) time for terminating recognition of the utterance based on utterance data stored in the memory, and terminate the recognition of the utterance when the utterance is not additionally input within the EPD time. The utterance data may include sentence data, which includes a pattern related to a sentence type of the utterance, and word data, which includes a word of the utterance.

The application relates to the technical field of semiconductor detection, and discloses an end-point detection device capable of preventing free jet flow and entrainment and a dynamic adjustment method, wherein the detection device comprises a light source module, a gas chamber module, a variable-diameter module, a segmented temperature control module and a detector module; the gas chamber module comprises coaxially nested first, second and third cylinder bodies, a first chamber is formed between the first and second cylinder bodies, the inner cavity of the third cylinder body is a second chamber, a third chamber is formed between the third and second cylinder bodies, the inner wall of the first cylinder body and the outer wall of the second cylinder body are provided with wall-attached flow guide structures, and the third cylinder body is densely provided with exhaust holes; the variable-diameter module is used for adjusting the flow area of an exhaust pipe; and the segmented temperature control module is used for forming a temperature gradient which decreases along the gas flow direction. Through the cooperation of the anti-entrainment gas chamber structure, the variable-diameter module and the segmented temperature control module, the jet flow entrainment effect can be inhibited, and the accuracy of gas concentration measurement and the reliability of clean process monitoring can be improved.

The embodiment of the present application provides a process step endpoint detection method and a semiconductor process equipment, the method comprises the following steps: in the process of executing a process step, obtaining an executed process time and an expected process time of the process step; according to the expected process time and the executed process time, calculating a delay end detection time of the process step; according to the delay end detection time, triggering a first timing monitor to start timing monitoring, and triggering a skip processing for the process step when the timing time of the first timing monitor reaches the delay end detection time. The flexible variation of the process step endpoint time can be realized according to the expected process time set by the user, and the actual execution time of the process step can meet the expected process time of the user; when the timing time of the timing monitor reaches the delay end detection time, the skip can be automatically executed, and the button of manual end detection does not need to be manually triggered, so that the automation is higher.

Techniques are described herein for a method of obtaining a token based on a conversation in real time. The method further includes predicting, using a large language model (LLM) and the token, a next token. The method further includes predicting, using a classifier and the next token, a completion of a user turn. The method further includes triggering a next turn of the conversation in real time using the completion of the user turn.

A computer-implemented method for generating synthetic data describing a sample may include: processing, by a system operatively coupled to a processor, sample information describing a set of original features of the sample, thereby producing processed model input information; and generating composite image data by analyzing a model and based on the processed model input information, the composite image data being usable to generate a composite image and describing at least a portion of the set of original features. The computer-implemented method may also include generating, by the system, a composite image of dummy samples based on the composite image data, the composite image having at least one dummy feature different from and based on the set of original features.

Systems and methods of persistent file protection integrate cloud-based backups with EDR file protection. Use of local persistent file cache (PFC) data with cloud-stored file segments, facilitates reconstruction of file states from multiple sources. File storage and retrieval efficiency are optimized, even when local session caches are invalidated or when extensive file modifications exceed storage quotas.

The present invention discloses an intelligent speech recognition method, device and related equipment based on classification identification. The method includes performing endpoint detection and feature extraction on raw speech data to obtain speech feature data; identifying the industry attributes of the raw speech data by the industry attribute identifier attached when the raw speech data is input; identifying the speech feature data through a trained general speech recognition model to obtain a recognition result; judging whether the confidence level in the recognition result is lower than a preset value; inputting the corresponding text with a confidence level lower than the preset value into the corresponding industry vertical module for optimized recognition; and selecting the text corresponding to the maximum confidence level as the final recognition result based on the confidence level in the recognition result and the optimized confidence level. By attaching the industry attribute identifier when inputting the raw speech data, it is convenient to subsequently call the corresponding industry vertical module for optimized recognition; no manual labeling is required, and a high accuracy rate is maintained by secondary verification and recognition correction.

This application provides a wafer processing system, an endpoint detection device, a method, and an electronic device. The wafer processing system includes: a wafer polishing assembly, including a polishing disk; an optical assembly disposed within the polishing disk and rotating synchronously with the polishing disk, used to acquire the reflection spectral signal of the wafer film layer in real time during the wafer polishing process; and a controller, communicatively connected to the optical assembly, used to select a matching spectral data processing strategy based on the current wafer polishing parameters to identify whether the reflection spectral signal is an abnormal spectrum; if not, to determine the current film layer thickness based on the reflection spectral signal; and to perform polishing endpoint determination based on the film layer thickness. The solution of this application can avoid misjudgment of the polishing endpoint caused by abnormal spectral data, ensuring the consistency and yield of wafer processing.

Disclosed is a method and device for image-based malware detection, and artificial intelligence-based endpoint detection and response system using same. The method includes the steps of: reading, as bytes of an unsigned integer of constant bits, an executable file or at least a portion of the code of an executable file located in a file path of a generated file obtained from a file generation event log detected in an EDR system or a processor, a memory, or a storage device associated with an endpoint, and converting same into a byte array; reading two bytes at a time from the byte array; using the two bytes of data as coordinates of an image of a preset size to increase a corresponding coordinate value, and adding an RGB channel to store an image array; and performing deep learning analysis using an image deep learning model on the image array.