675results about How to "Narrow down the search space" patented technology

An enhanced system for speech interpretation is provided. The system may include receiving a user verbalization and generating one or more preliminary interpretations of the verbalization by identifying one or more phonemes in the verbalization. An acoustic grammar may be used to map the phonemes to syllables or words, and the acoustic grammar may include one or more linking elements to reduce a search space associated with the grammar. The preliminary interpretations may be subject to various post-processing techniques to sharpen accuracy of the preliminary interpretation. A heuristic model may assign weights to various parameters based on a context, a user profile, or other domain knowledge. A probable interpretation may be identified based on a confidence score for each of a set of candidate interpretations generated by the heuristic model. The model may be augmented or updated based on various information associated with the interpretation of the verbalization.

Improved timing synchronization and access control techniques for use in an orthogonal frequency division multiplexed (OFDM) wireless system or other type of wireless communication system. In accordance with the invention, an uplink synchronization and access control system is provided in which mobile stations transmit certain timing and access signals in dedicated intervals in an uplink stream. Access control is illustratively implemented as a two-stage process in which a given mobile first transmits a generic uplink access signal in one of the intervals. If this access is accepted, the base station transmits an access acknowledgment containing initial timing and power corrections, along with initial channel assignments on which the mobile can initiate a call set-up process. For re-synchronization, mobiles transmit timing synchronization signals in the dedicated timing and access intervals. The base station measures the arrival time of the signals, and sends back appropriate timing corrections. The invention thereby ensures that orthogonality between mobiles is maintained.

Integrated intelligent system adapted for any operating system and / or multi-OS computing environment seamlessly having sensory input / data acquisition cum recording server group and / or analytics server group enabling fail-safe integration and / or optimized utilization of various sensory inputs for various utility applications. Also disclosed as added advancements include intelligent method / system for cost-effective and efficient band adaptive transferring / recording sensory data from single or multiple data sources to network accessible storage devices, fail safe and self sufficient server group based method for sensory input recording and live streaming in a multi-server environment, intelligent and unified method of colour coherent object analysis, face detection in video images and the like, resource allocation for analytical processing involving multi channel environment, multi channel join-split mechanism adapted for low and / or variable bandwidth network link, enhanced multi-colour and / or mono-colour object tracking and also an intelligent automated traffic enforcement system.

The invention relates to a flight path planning method based on a sparse A* algorithm and a genetic algorithm and belongs to the technical field of flight path planning of an unmanned aerial vehicle (UAV). According to the characteristics of flight path planning, the method comprises the following steps: planning an initial reference flight path by the utilization of the sparse A*search (SAS) algorithm, wherein constraint conditions are combined into algorithm search, so that useless nodes in a search space can be effectively deleted and the search time is shortened; and when emergent threat occurs during real-time flight of the UAV, performing real-time flight path planning by the utilization of the genetic algorithm and generating a flight path with local optimum or approximate optimum until the threat disappears and the UAV returns the original global optimum reference flight path and continues flying. The method provided by the invention has high real-time performance and rapidity; the searched flight path is closer to the actual UAV optimal flight path; and the method can be applied to the technical fields of robot path planning, urban vehicle path planning and the like under complex environments.

The data constraint framework solution of the present invention addresses data quality issues by standardizing, verifying, matching, consolidating and merging data records using powerful inexact matching logic and search reduction technologies. The data conditioning framework uses these technologies to more efficiently condition data to improve the quality of data and / or resolve quality data issues such as incomplete, inaccurate and duplicate data records. For example, the data conditioning framework is used to “cleanse” incorrect, incomplete and duplicate data from a data source, such as an information system. The data conditioning framework uses the following approximate searching and matching techniques to improve the efficiency of the approximate matching, reduce the search space for approximate matching, and improve the speed of executing approximate searches and matches: 1) inexact trimmed matching, 2) adaptive search ordering, 3) cascading search space reduction, 4) tiered and metric indexing, and 5) domain knowledge matching.

The invention discloses a method for detecting code similarity based on semantic analysis of a program source code, which relates to computer program analyzing technology and a method for detecting complex codes of computer software. The method solves the prior problems of low similarity detection accuracy and high computing complexity on the codes of different syntactic representations and similar semantemes, and also solves the problem of incapability of realizing large-scale program code similarity detection. The method comprises the following steps: resolving two segments of source codes to be detected into two control dependence trees of a system dependence graph respectively and executing basic code standardization respectively; utilizing a measure method to extract candidate similar code control dependence trees of the control dependence trees which are subjected to the basic code standardization; executing an advanced code standardization operation on extracted candidate similar codes; and computing semantic similarity to obtain a similarity result so as to finish the code similarity detection. The method is applied to source code piracy detection, software component library query, software defect detection, program comprehension and the like.

Methods and apparatus for reducing the search space processed by mask matching methods. The search space is reduced by grouping the candidate bit patterns into groups and subgroups that have internal bit agreement between the members. By only applying the mask matching methods to a select number of groups selected by their bit agreement with the target bit pattern, the computation time and memory requirement of the mask matching method is reduced.

A system, process and software arrangement are provided to determining data associated with at least one structural change of tissue. In particular, a first optical coherence tomography (“OCT”) signal which contains first information regarding the tissue at a first stress level, and a second OCT signal which contains second information regarding the tissue at a second stress level are received. The first and second information are compared to produce comparison information. The data associated with the at least one structural change is determined as a function of the comparison information and further information associated with (i) at least one known characteristics of the tissue and / or (ii) characteristics of an OCT system. Further, at least one optical coherence tomography (“OCT”) signal which contains information regarding the tissue can be received, and the modulus of the tissue may be determined as a function of the received at least one OCT signal.

The invention discloses a cascade reservoir multi-objective optimization scheduling method based on an improved artificial bee colony algorithm. The method comprises the following steps that s11, the basic information data of a cascade reservoir system are acquired; s12, a multi-objective scheduling model including power generation, estuarine ecology and water supply is established according to the information of the reservoir system; and s13, the optimal scheduling scheme of the cascade reservoir system is solved by performing the improved artificial bee colony algorithm. Global optimization of the reservoir scheduling problem can be realized so that the calculation efficiency and accuracy can be enhanced, and a new approach can be provided for solving the multi-objective optimization scheduling problem of the cascade reservoir system.

The invention relates to the technical field of face recognition, in particular to a rapid detection method of a face in a high-resolution color image under a complex background. The invention comprises the following steps: building a face skin color mixed model which is composed of two color space restrictions of RGB and YCbCr according to a large amount of acquired skin color sample data to determine the skin color pixel; rapidly skipping non-face regions by adopting a whole skin color pixel ratio to improve the location efficiency of a face candidate region; then using the improved face rapid detection algorithm which is based on forward characteristic selection to realize face preliminary judging of the face candidate region; and finally utilizing a false alarm restraining method which is based on space restriction and geometric restriction to further lower false detection rate and complete face detection. The invention can realize rapid location of a plurality of frontal faces in images at a high detection rate under the condition of low false alarm rate; the good performances of the invention are proved by the results of the test set of Bao open database and a plurality of video frequencies and high resolution color images.

The invention provides a multi-target route planning and united searching method of a drone in an urban low-altitude environment, and belongs to the technical field of drone route planning. The method comprises steps of firstly, based on a static-state known obstacle, constructing a static-state safe index map, and based on an obstacle which is actively detected by the drone in a flight process and is not marked in a geographical map, constructing a dynamic state safe index map on an airborne platform; and then, based on offline searching and online searching, solving and optimizing a multi-target route planning problem of flight time and flight safety of the drone in an united manner. According to the invention, a path costing little from a start point to a termination point is planned in an offline manner; when an unknown obstacle is detected, a changed flight track is planned in an online manner through online searching, so a dynamic obstacle is avoided; and there is a quite small searching space in the online searching, so a safe route can be planned again from the drone rapidly, so requirements on real-time planning of routes for drones are met.

A system and method for detecting speech utilizing audio and video inputs. In one aspect, the invention collects audio data generated from a microphone device. In another aspect, the invention collects video data and processes the data to determine a mouth location for a given speaker. The audio and video are inputted into a time-delay neural network that processes the data to determine which target is speaking. The neural network processing is based upon a correlation to detected mouth movement from the video data and audio sounds detected by the microphone.

The invention discloses a deep learning-based machine reading understanding method, system and device, and a medium. The method comprises the steps of constructing a question classification model, a support sentence search model and an answer determination model; inputting the target problem into a problem classification model, predicting through the problem classification model, and outputting toobtain a target problem type; inputting the target question and a to-be-read understanding article into a support sentence search model, predicting through the support sentence search model, and outputting to obtain a support sentence sequence related to the answer of the target question; and inputting the target question, the answer type corresponding to the target question type and the supportsentence sequence related to the answer of the target question into an answer determination model, performing prediction through the answer determination model, and performing output to obtain the answer of the target question. The method is suitable for English machine reading understanding tasks, can effectively process the situation that many article paragraphs exist and answers can be obtainedonly by reasoning in multiple paragraphs, and improves the accuracy of machine reading understanding.

The invention discloses an efficient verifiable multi-keyword sequencing searchable encryption method supporting a preference search and a logic search. The encryption method comprises the following steps that a data owner encrypts a document, constructs a cryptogram index based on a document collection, generates summary information of the document by utilizing a secret key and the content of thedocument, and then sends a cryptogram document, an encrypted index and the summary information to a cloud server; a data user shares the information of the secret key, which is generated by the dataowner, generates a query trap door according to a query, and sends an encrypted query trap door and the number K of to-be-obtained documents to the cloud server; the cloud server can execute safe inner product operation after receiving the cryptogram index and the query trap door, searches for K documents most relevant to the query of the user, sequences according to relevant values among the K documents and the query, then generates verification objects, and finally returns the most relevant K documents and the verification objects to the data user, and the data user verifies the correctnessand completeness of a return result through a verification algorithm.

The invention discloses a reinforcement dual-channel sequence learning-based dialog reply generation method and system. The method comprises the following steps of: (1) modeling a context to obtain acontext semantic vector; (2) carrying out combined learning on a current dialog and the context semantic vector by utilizing an encoder so as to current a current dialog vector and an encoder vector;(3) inputting the context semantic vector and the current dialog vector into a decoder so as to obtain a first channel dialog reply draft and a decoder vector; (4) inputting the encoder vector, the decoder vector and the first cannel dialog reply draft into an embellishing device to carry out embellishing, so as to generate an embellished dialog reply of a second channel; (5) optimizing a target function by utilizing a reinforcement learning algorithm; and (6) ending model training and generating and outputting a dialog reply. By utilizing the method and system, dialog generation models can grasp global information more deeply, and replies more according with dialog scenes and having substantial contents can be generated.

A method of designing a network circuit for a telecommunications network is disclosed. The method includes storing first model data in a database representing a first model of the telecommunications network. Second model data is received representing a second model of the telecommunications network. The second model data defines node groups, each representing a plurality of network nodes of the first network model, and interconnections between node groups, each representing connectivity between node groups in the first network model. The method involves searching the second model data to select a route from a start node group to an end node group, the route specifying a plurality of node groups connected by interconnections in the second network model; and searching the first model data to select a circuit from a node of the start node group to a node of the end node group.

The embodiment of the invention discloses a route planning method and device. The method comprises the steps of acquiring a route planning request containing a start point and an end point; acquiring an experience weight set of a section between the start point and the end point according to the route planning request, wherein the experience weight set is determined according to historical routes of a user; according to the experience weight set of the section, performing route planning for the route planning request by adopting a heuristic search algorithm. According to the technical scheme provided by the embodiment of the invention, an existing route planning technology can be optimized, and planning speed and planning result accuracy can be improved.

The invention discloses a method for realizing high-precision location based on Big Dipper system civil carrier phase combination. The method comprises the steps of calculating the original user position and utilizing the original user position to calculate the GEO superwide lane whole circle ambiguity; taking the integer of the observed quantity of GEO superwide lane combination, calculating the GEO wide lane whole circle ambiguity (img file='DDA00002471851000011.TIF' wi='160' he'='47' / ), calculating the error of double difference observed quantity of the GEO superwide lane combination, and determining the searching range of the ambiguity of a GEO wide lane; conducting dimensionality reduction search on the GEO wide lane whole circle ambiguity (img file='DDA00002471851000012.TIF' wi='134' he'='47' / ) to obtain the accurate value of the GEO wide lane whole circle ambiguity; selecting an MEO / IGSO satellite to initially calculate the wide lane whole circle ambiguity; utilizing the ambiguity combined relationship between a superwide lane and a wide lane to obtain the double difference ambiguity on a B1 frequency point, a B2 frequency point and an S frequency point; utilizing the sum of the double difference ambiguity on the B1 frequency point, the B2 frequency point and the S frequency point to receive carrier phase of machine measurement and conduct planet base double difference location, and calculating the high precision position of a user. Through the method, the purpose of quick high precision location of Big Dipper carrier phases is achieved.

A face recognition neural network generates small dimensional facial feature vectors for faces in received images. A multi-task learning classifier network receives generated vectors and outputs classifications for the corresponding images. The classifications are built by respective sub-networks each comprising classification layers arranged in levels. The layers in at least some levels receive as input intermediate outputs from an immediately upstream classification layer in the same sub-network and in each other sub-network. After training, a resultant facial feature vector corresponding to an input image is received from the neural network. Resultant classifiers for the input image are received from the classifier network. A database is searched for a subset of reference facial feature vectors having associated classification features matching the resultant classifiers. Vector matching identifies which one(s) of the reference facial feature vectors in the subset of reference facial feature vectors most closely match(es) the resultant facial feature vector.

The invention provides a method for improving resolving success rate of medium and long baseline GPS integral cycle fuzziness, and aims to provide a method with advantages of high fuzziness real solution precision and high success rate. The method is realized through a technical solution which is characterized in that the method comprises the steps of setting an elevation mask angle by means of monitoring and measuring software, acquiring rough coordinates of a main measurement station and an auxiliary measurement station, and extracting useful observation data; selecting a reference satellite, comparing observation data of the main measurement state with the observation data of the auxiliary measurement station, establishing a double-difference observation equation by means of the observation data of the main measurement station and the auxiliary measurement station, performing state updating and parameter estimation on an observed equation by means of a Kalman filter, and solving a fuzziness float solution; then, converting the fuzziness float solutions of GPS double-difference carrier L1 and L2, and forming a vector set which comprises a wide lane integral cycle fuzziness and a carrier L1 fuzziness; searching the wide lane integral cycle fuzziness by means of a LAMBDA algorithm, obtaining the integer solution of the wide lane integral cycle fuzziness after fixation, and finally determining the success rate of the resolved fuzziness through a preset threshold, and confirming a correctly resolved integral cycle fuzziness.

A method for acquiring the motion information of coronary artery, which belongs to the field of medical detection technology, is used for solving the problem in acquiring the motion information of heart. The technical scheme is that a 3D motion estimation method based on elastic registration is adopted after achieving 3D reconstruction of branch skeleton of main blood vessels at each moment in X-ray coronary artery angiogram sequences; the motion of blood skeleton is transformed to be matched with skeleton lines at continuous time moments; the motion vector and the motion trajectory of each blood skeleton point in cardiac cycle are calculated; and the motion information including global and the local motion parameters of the coronary artery in the cardiac cycle is extracted according to the estimated motion vector of each blood skeleton in the image sequence. The method has the advantages of high accuracy in extracting the motion information of blood vessels, convenient operation and high work efficiency.

The invention provides a high water-cut oilfield well position determining method based on an evolutionary algorithm. The method comprises the steps of building an oil reservoir model of a predetermined oil area, and dividing the oil reservoir model into grids; obtaining the oil reservoir data, and obtaining a surplus oil potential enrichment area in a history matching mode according to the oil reservoir data; computing a well position constraint value for representing the surplus oil distribution and injected water waves and volume in the surplus oil potential enrichment area to obtain the well position advantage area; carrying out iteration on an oil well in the advantage area through the evolutionary algorithm to compute a fitness function, and evaluating the fitness function through a Kriging algorithm, wherein the evolutionary algorithm is adjusted according to the fitness function, and the Kriging algorithm considers the multi-dimensional space, the nugget effect and the local weight; obtaining a maximum fitness function value according to the iteration computing of the evolutionary algorithm so as to determine the well position. The improved evolutionary algorithm and the Kringing algorithm serve as the optimization tools, a dynamic potential index is built to reduce the searching space of the evolutionary algorithm, and the optimization computing convergence precision and speed are greatly improved.

The invention discloses a method for positioning an iris and a pupil based on eye structure classification. By the method, a face is detected in an image acquired by a system, and an eye image is extracted. Under a remote non-intrusive condition, the acquired eye image can be changed within an extremely large range, for example, the radius of the iris can be changed within dozens to hundreds of pixels. According to the method, before an iris boundary and a pupil boundary are precisely positioned, the eye image is automatically subjected to structural classification by an unsupervised learning technology; due to the classification, the iris and the pupil can be roughly positioned; the size of an iris region can be effectively estimated; stray data of a non iris boundary and a non pupil boundary are greatly eliminated; and a search space aiming at positions and sizes of the iris and the pupil is reduced. In the reduced search space, constrained optimization is performed according to fixed features of the iris and the pupil; and the optimal iris boundary and the optimal pupil boundary are searched. By the method, the stability and the precision of positioning of the iris and the pupil can be greatly improved; and the method is particularly suitable for a remote non-intrusive iris obtaining system.

The invention provides a multi-unmanned aerial vehicle four-dimensional track collaborative planning method and system. The method comprises the following steps: giving unmanned aerial vehicle three-dimensional task space and task time range, take-off sites and destination sites of multiple unmanned aerial vehicles, and flight time constraint information; discretizing the three-dimensional task space into three-dimensional cuboid grids, numbering the three-dimensional cuboid grids, marking the position of the terrain and the position of the threat space in the three-dimensional task space, and abstracting the three-dimensional task space into a directed graph; constructing a single-target multi-unmanned aerial vehicle four-dimensional flight path planning model according to the unmanned aerial vehicle three-dimensional task space and the task time range; and solving the single-target multi-unmanned aerial vehicle four-dimensional route planning model to obtain a planned route, take-off time and flight speed of each unmanned aerial vehicle. Through time optimization, the problem that in a traditional method, only a three-dimensional path is optimized, but time optimization is ignored is solved, the flight path time of the unmanned aerial vehicle is determined better while path planning is carried out, and the utilization rate of space-time resources is increased.

The invention relates to a forecasting method for a multi-stage differential evolution protein structure based on abstract bulge estimation. The method comprises the following steps: firstly, calculating the distance from each conformation individual in a current colony to a new conformation and performing ascending sorting according to the distance; then selecting the part of the new conformation individual close to a abstract bulge lower-limit estimation support surface of the conformation individual, thereby acquiring an energy lower-limit estimation value of the new conformation individual; calculating an average estimation error between the energy lower-limit estimation value of all the new conformation individuals and a practical energy value; dividing the whole algorithm into a plurality of optimizing stages according to the change in the average estimation error; judging the stage of the present iteration according to the average estimation error in the last iteration; and designing different strategies for all the stages and generating the new conformation individual. The forecasting method for the multi-stage differential evolution protein structure based on the colony abstract bulge estimation provided by the invention is high in forecasting precision and low in calculation cost.

Approaches based on dynamic tainting to assist transform users in debugging input models. The approach instruments the transform code to associate taint marks with the input-model elements, and propagate the marks to the output text. The taint marks identify the input-model elements that either contribute to an output string, or cause potentially incorrect paths to be executed through the transform, which results in an incorrect or a missing string in the output. This approach can significantly reduce the fault search space and, in many cases, precisely identify the input-model faults. By way of a significant advantage, the approach automates, with a high degree of accuracy, a debugging task that can be tedious to perform manually.

The invention discloses a method for quickly planning and mixing paths on the basis of A-star algorithms.The method includes steps of S1, dividing a target map into a plurality of map blocks with equal sizes; computing and storing the shortest paths between all adjacent points and nodes in each map block; S2, judging affected road segments in the map blocks affected by events when the events occur, computing the shortest paths among all the adjacent points and the nodes in each map block related to the affected road segments after the affected road segments are excluded; S3, selecting the various map blocks, where paths from start points to end points pass, on the basis of the A-star algorithms, and connecting the shortest paths in the various map blocks with one another after the affected segments are excluded so as to form a plurality of alternative paths.The adjacent points represent intersection points of boundaries of roads and the map blocks, and the nodes are the start points or cross points of the roads.The alternative paths with the minimum time cost are the optimal paths.

The invention relates to a face and iris image fusion and recognition method based on a hierarchical structure, which comprises a registration step R and a recognition step S, wherein in the registration step R, after a feature extraction link R1 for face and iris images of a registered user is carried out, the images are stored in a registration database; and in the recognition step S, the face and iris images are used for recognizing and searching on the basis of the given registration database. The method of the invention improves the accuracy, the usability and the searching efficiency of a biological feature recognition system and has excellent expansibility and compatibility. In the method, the biological feature recognition system based on face and iris fusion can effectively recognize remote pedestrians and has wide application prospects in airports, stations, banks and other safe occasions.

A property vector representing extractable measurable properties, such as musical properties, of a file is mapped to semantic properties for the file. This is achieved by using artificial neural networks “ANNs” in which weights and biases are trained to align a distance dissimilarity measure in property space for pairwise comparative files back towards a corresponding semantic distance dissimilarity measure in semantic space for those same files. The result is that, once optimised, the ANNs can process any file, parsed with those properties, to identify other files sharing common traits reflective of emotional perception, thereby rendering a more liable and true-to-life result of similarity / dissimilarity. This contrasts with simply training a neural network to consider extractable measurable properties that, in isolation, do not provide a reliable contextual relationship into the real-world.