14 results about "End point detection" patented technology

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.

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.

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.

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.

A new approach is proposed to support integration of EDR data from a plurality of EDR tools / sources into an Open XDR framework in an automated manner. First, EDR data generated by each of the plurality of EDR tools covering a plurality of assets is ingested into the Open XDR framework. The ingested EDR data is then normalized through a unified EDR data model. The normalized EDR data is further enriched with one or more new data fields to better correlate the EDR normalized data from the plurality of EDR tools. A plurality of alerts are then generated from the normalized and enriched data along one or more alert pathways to improve fidelity of the plurality of alerts. The plurality of alerts are correlated with the contextual information of the plurality of assets as well as information from other data sources to identify a set of incidents of suspicious activities.

The system and method for performing cybersecurity threat detection on a resource in a cloud computing environment are presented. The method includes providing a sensor to the resource; configuring the sensor to detect an event in the resource from a data link layer communication; matching the event to a rule, the rule specifying a mitigation action if a condition is met; initiating the mitigation action in accordance with the rule; and including data of the event in a software bill of materials (SBOM).

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.

This invention discloses a VDR speech endpoint detection method, comprising: extracting feature information from an audio signal, obtaining the first-order and second-order differences of the four feature information respectively; inputting the zero-padded feature map into a residual network with an attention mechanism to extract complex abstract features of the feature map; calculating the feature centroids corresponding to the initial output values ​​of 0 and 1; searching for abrupt changes in the initial output of speech endpoint detection with a duration of less than 100ms, defining them as short-term abrupt changes, and calculating the similarity of the feature centroids of the abrupt changes with the feature centroids of the 0 and 1 judgment results of the entire audio file respectively; updating the VDR speech endpoint detection output value based on the feature centroid similarity estimation results of the short-term abrupt changes, and obtaining the final VDR speech endpoint detection output value. This method avoids short-term endpoint detection abrupt changes, thereby accurately locating the speech position in the VDR audio signal.

The presently disclosed subject matter includes a system and method for decentralized endpoint detection and response (EDR) that enables scalable, relevance-driven security monitoring across large endpoint populations. Each endpoint locally processes security queries and generates a compact Lightweight Local Response (LLR) indicating whether a local result (LR) is relevant. Full LRs are transmitted only when their associated LLRs are confirmed by the system, reducing bandwidth, storage, and processing overhead. A dynamic batch cycle mechanism controls query distribution, allowing some queries to scale broadly while directing others to selected endpoints based on metadata. Query execution continues across batches until a batch termination condition (BTC) is met, followed by evaluation of a query termination condition (QTC) based on real-time response characteristics. This decentralized approach eliminates the need for centralized telemetry aggregation, improves scalability, enhances privacy, and reduces operational costs.

A method includes providing first OES time trace data from a substrate processing operation to a trained machine learning model. The first OES time trace data has a first set of wavelengths. The method further includes obtaining, from the trained machine learning model, synthetic OES time trace data for the substrate processing operation, the synthetic OES time trace data having a second set of wavelengths different from the first set. The method further includes obtaining second OES time trace data from the substrate processing operation having the second set of wavelengths. The method further includes determining a process endpoint for the substrate processing operation based on the synthetic OES time trace data and the second OES time trace data. The method further includes performing an action in accordance with the process endpoint.

This application discloses a speech segmentation method, apparatus, device, and readable storage medium, relating to the field of speech interaction technology. It includes: performing real-time semantic analysis on the user's speech stream locally on the terminal device to determine the semantic integrity status, and dynamically selecting a target threshold from a preset set of silence thresholds to extract audio segments and send them to the cloud; the cloud performs endpoint detection on the audio segments, obtains the start and end timestamps of the speech, and calculates the time interval between the current segment and the previous segment; when the interval is less than or equal to a preset listening window, it performs a semantic correlation judgment on the two segments based on rules and a semantic model, and decides whether to splice them. This application solves the problems of response delay and semantic confusion caused by fixed thresholds through semantically driven dynamic threshold adjustment and cloud-based semantic-level splicing decision-making, thereby improving the response speed and recognition accuracy of speech interaction.

The application provides a nondestructive detection method and program product for watermelon maturity, which can be applied to the field of nondestructive detection of agricultural products. The method comprises the following steps: applying at least one preset action to a watermelon to be detected, and obtaining an original soundprint signal generated by the watermelon to be detected; performing endpoint detection on the original soundprint signal to extract an effective soundprint signal, wherein the effective soundprint signal is used to represent a signal containing the acoustic response characteristics of the watermelon to be detected; performing feature extraction based on the effective soundprint signal to obtain acoustic features and time-frequency domain features; performing feature fusion on the acoustic features and the time-frequency domain features to obtain fused features; and inputting the fused features into a pre-trained maturity recognition model to obtain a maturity detection result of the watermelon to be detected, wherein the maturity recognition model is constructed based on a deep neural network.

This application provides an eddy current detection method, an endpoint detection system, and a chemical mechanical polishing (CMP) device. The eddy current detection method includes: during relative motion between a wafer and a polishing pad, monitoring eddy current signals using an eddy current sensor mounted on a polishing pad. The eddy current signals include: a marker signal and a wafer morphology feature signal; determining the signal angle of the marker based on the marker signal, and determining the angular deviation value between the signal angle and the actual angle of the marker; calibrating the wafer morphology feature signal based on the angular deviation value; wherein, during wafer polishing, the eddy current sensor forms a circular trajectory as the polishing pad rotates, a first line connecting the horizontal projection of the polishing pad center and the horizontal projection of the wafer center intersects the horizontal projection of the circular trajectory, and the horizontal projection of a preset marker intersects the horizontal projection of the circular trajectory, while the position of the marker remains unchanged. This application can improve the accuracy of determining wafer thin film thickness.

The present disclosure provides a data processing method, system, electronic device and computer readable storage medium, the method comprising: in response to a data processing request of a resource control terminal, obtaining audio data requested to be processed by the resource control terminal; sending a start message to an audio processing device, the start message being used to instruct the audio processing device to start a voice endpoint detection function; receiving first confirmation information returned by the audio processing device, the first confirmation information being used to indicate a started state of the voice endpoint detection function; in response to the first confirmation information, sending the audio data to the audio processing device in real time; and receiving an audio processing result sent by the audio processing device. According to the embodiments of the present disclosure, the processing performance of the audio data can be improved.