609 results about "Benchmark data" patented technology

This invention is a system, method, and apparatus for detecting compromise of IP devices that make up an IP-based network. One embodiment is a method for detecting and alerting on the following conditions: (1) Denial of Service Attack; (2) Unauthorized Usage Attack; and (3) Spoofing Attack. A survey of services running on the IP device, historical benchmark data, and traceroute information may be used to detect a possible Denial of Service Attack. A detailed log analysis and a passive DNS compromise system may be used to detect a possible unauthorized usage. Finally, a fingerprint of the IP device or its configuration settings, a watermark inserted in the data-stream, and a private key burned into the IP devices' physical memory may be used to detect a possible spoofing attack. The present invention may be used to help mitigate intrusions and vulnerabilities in IP networks.

Personality, behavior and competency assessment data is obtained for a sales person (or group of sales persons), particularly via a computer-implemented interface. The resulting overall individual (or aggregated) assessment data is then compared with high sales performer benchmark data to identify performance improvement opportunities that, in turn, are used to identify at least one specific action for the sales person (or group of sales persons). Such techniques may be use with sales management personnel as well. Reports based on the assessment results and comparisons thereof with high sales performer benchmark data may be created and provided to the individuals being assessed or persons associated with the individual sales persons. The personality, behavior and competency assessment data may also be used to identify predictive characteristics of high sales performers that may be subsequently used to predict other characteristics of individuals or groups.

Presented herein are embodiments of a multimodal Recurrent Neural Network (m-RNN) model for generating novel image captions. In embodiments, it directly models the probability distribution of generating a word given a previous word or words and an image, and image captions are generated according to this distribution. In embodiments, the model comprises two sub-networks: a deep recurrent neural network for sentences and a deep convolutional network for images. In embodiments, these two sub-networks interact with each other in a multimodal layer to form the whole m-RNN model. The effectiveness of an embodiment of model was validated on four benchmark datasets, and it outperformed the state-of-the-art methods. In embodiments, the m-RNN model may also be applied to retrieval tasks for retrieving images or captions.

Systems and methods for estimating power consumption in a network of computing devices are described. Operational information of a target server is periodically received and compared to benchmark data of a model of the target server. The operational information comprises performance data of the target server during a predefined time interval. Power consumption of the target server is estimated using the performance and benchmark data. The benchmark data is recalibrated if an error in the estimated power consumption is detected. An agent installed on the target server for collecting performance data is described. The target server can be a virtualized server, in which case, the agent acquires at least some of the performance data from a hypervisor of a physical server that hosts the target server.

Systems and methods are disclosed for providing benchmark services to customers. The services may be provided through a benchmark provider. The benchmark provider may build a benchmark database based on data collected from customers and / or third party sources. The submission of benchmark data can be made optional to the customer and, thus, not a requirement to receive service(s) from the benchmark provider. In addition, the benchmark database may be built as a comprehensive database, such that it is well-populated and provides benchmark data for many or all types of functional areas (e.g., FIN, CRM, SCM, SRM, HR, etc.), industries and / or sizes of companies. Offers for benchmark services may be presented to customers and specific service(s) may be provided in accordance with the service election(s) made by each customer.

Methods and systems are provided for processing and storing data objects by means of one or more processes running in a computer system. The computer system may have one or more electronic data structures. The electronic data structures may include a benchmark definition and may be formatted to be readable by a SGML-based language. In one implementation, the methods and systems present one or more of the electronic data structures to second parties by electronic means.

The Invention relates to an electronic data structure comprising benchmark data, said electronic data structure being a file having a format readable by a SGML-based language. The invention further relates to a method for processing and storing data objects by means of one or more processes running in a computer system having one or more of the electronic data structures comprising a step of incorporating one or more of the benchmark data or benchmark definition contained in said one or more electronic data structures into data structure or into a software for support of business processes.

This invention is a system, method, and apparatus for detecting compromise of IP devices that make up an IP-based network. One embodiment is a method for detecting and alerting on the following conditions: (1) Denial of Service Attack; (2) Unauthorized Usage Attack (for an IP camera, unauthorized person seeing a camera image); and (3) Spoofing Attack (for an IP camera, unauthorized person seeing substitute images). A survey of services running on the IP device, historical benchmark data, and traceroute information may be used to detect a possible Denial of Service Attack. A detailed log analysis and a passive DNS compromise system may be used to detect a possible unauthorized usage. Finally, a fingerprint (a hash of device configuration data) may be used as a private key to detect a possible spoofing attack. The present invention may be used to help mitigate intrusions and vulnerabilities in IP networks.

The invention discloses a method for detecting refrigerant leakage of an air conditioning system and the air conditioning system. The method comprises the steps of: S10, collecting the state parameters of the air conditioning system when the external environment temperature is T in an operating process of the air conditioning system and storing the collected data at the temperature for the first time as reference data; S20, in the operating process of the air conditioning system, when the external environment temperature reaches T again, collecting the state parameters of the air conditioning system again; S30, comparing the collected data with the reference data and calculating deviation therebetween; and S40, when the deviation exceeds a predetermined value, judging the condition of refrigerant leakage. When the method disclosed by the invention is used for detecting refrigerant leakage, the detection result is not affected by an actual installing condition; the influence on the environment temperature can be eliminated, so that the extent of refrigerant deficiency in the air conditioning system can be accurately judged, and the detection result is more reliable. The invention further discloses the air conditioning system.

Systems and methods of performing comparisons of client data and benchmark data are described. When client data is compared to internal benchmark data, comparative result values may be computed by performing a multidimensional operation with respect to client data for a particular dimension member and client data for additional dimension members that are related to the particular dimension member. When client data is compared to external benchmark data, a client result value may be computed based on a multidimensional operation and a benchmark result value may be determined based on a mapping of client dimension members to benchmark dimension members. The result values may be provided by a server to a client, which may generate a graphical user interface (GUI) based on the result values.

The invention aims at providing a positioning method based on cloud computing. The method comprises steps: 1, a user terminal transmits reference data to the cloud end; 2, the cloud end carried out position computation according to the reference data; and 3, the cloud end returns position information. When the technical scheme of the invention is adopted, the problems that the computing ability of the user terminal is insufficient, the power consumption during the operation is large, the computing time is long, the real-time precision of a single machine is not high and the like in the current high-precision positioning structure and mode can be solved. The user only needs to pay for little network service fee and quick, high-precision and real-time positioning results can be acquired. The user terminal can be freed from heavy positioning computation, thereby reducing the power consumption of the user terminal and improving the continuous navigation ability. Once the cloud end is built successfully, the system can be used for a long time. Flexible use modes are provided. For the user terminal, especially the terminal user of Beidou, the cost and the power consumption can be greatly reduced, and industrialization development of Beidou is facilitated.

First, benchmark data are collected on two thin slices of an object to get measured k-space trajectories. To find delays on different physical gradients, the root mean square error between estimated and measured k-space trajectories is minimized for different delays. To reduce the error further, an eddy current compensation is introduced along each physical gradient axis. A more accurate estimation of the k-space trajectory can be obtained to image the object.

Turnkey maintenance of a customer's aircraft fleet is managed by a single management service provider (MSP) controlling integrated maintenance and materials services from a central operations site. The MSP converts data received directly from on-board aircraft systems into information it uses to manage maintenance service providers and parts suppliers. The MSP contracts with and manages maintenance, repair and overhaul organizations (MROs) who perform the maintenance on the customers' aircraft at line and base stations. The MSP either remotely manages part inventories at the customer's site, or manages suppliers who deliver the parts to the MROs. Maintenance planning, scheduling and execution information is exchanged between the MSP, MROs, part suppliers and the customers through a shared data communication network controlled by the MSP. The MSP charges the customer for the maintenance services based on a flat rate per unit of aircraft flying time.

A system and method may be used in generating security profiles for mobile device applications. An application may be installed and executed in a controlled execution environment. Physical, functional, data and / or communications parameters of the executing application may be monitored and / or analyzed. In this regard, the analysis may comprise evaluation based on comparison against benchmark data for the specific type of application being executed and for the specific execution environment. A security profile may be generated based on the monitoring and / or analysis. In this regard, the security profile may include security classification information for the application. The security profile may be made available to servers and / or to user mobile devices.

The invention discloses a method for improving accuracy of a geomagnetic parking stall detector. The method includes: by collecting sample data of a background magnetic field, calculating and recording benchmark data of the background magnetic field as reference standards; respectively determining calculated weight values in three axial directions according to stability of the sample data, collected by a triaxial magnetoresistive sensor, of the background magnetic field in three axial directions; turning off a wireless communication module when reading data of the triaxial magnetoresistive sensor, and storing the data into a buffer queue; calculating magnetic field changes for the data in the buffer queue, performing median filtering, and performing parking stall state determination according to changes of the magnetic field data; calibrating the benchmark data of the background magnetic field under the condition of free parking stalls; uploading the parking stall information and battery level information through the wireless communication module prior to entering the sleep state; waking through an RTC (real-time clock) in a timed manner, continuing executing from the step 3, and so on. Compared with the prior art, the method has the advantages that influence caused by temperature drift compensation and time drift of the sensor on parking stall determination accuracy can be restrained effectively, and the method is higher in adaptability, lower in cost and the like.

A system and method is disclosed for modeling, predicting, forecasting, and analyzing organizational spend / procurement, such as travel spend. In various embodiments, a method for modeling, predicting, budgeting, and optimizing business spend and procurement is provided comprising obtaining benchmark data, obtaining opportunity assessment data, obtaining organizational cost data, modeling the opportunity assessment data in comparison to the benchmark data, and determining a potential cost result based on the modeling. The method may further comprise adjusting the opportunity assessment data to achieve a desired cost result.

The invention provides a dynamic vehicle weighing method and system; all shafts of the vehicle to be weighed subsequently drive pass a weighing platform, wherein, the method includes the following steps: the shaft load signal and the shaft type signal of all the shafts of the vehicle to be weighed are collected; the dynamic weighed data and shaft type data of the shafts are generated according to the shaft load signal and the shaft type signal of the shafts; the dynamic weighed data is processed to obtain and weigh effective data; trigonometric function approximation processing is conducted on the shaft weight effective data to obtain the shaft weight benchmark data and the revised data; the shaft weight is obtained according to the shaft type data and the shaft weight benchmark data and the revised data. The invention is used to solve the problem of dynamic weighing of vehicles passing the weighing platform at medium and low speed, and provide weighing schemes meeting the weighing precision requirement for different vehicle types.

This invention is a system, method, and apparatus for detecting compromise of IP devices that make up an IP-based network. One embodiment is a method for detecting and alerting on the following conditions: (1) Denial of Service Attack; (2) Unauthorized Usage Attack (for an IP camera, unauthorized person seeing a camera image); and (3) Spoofing Attack (for an IP camera, unauthorized person seeing substitute images). A survey of services running on the IP device, historical benchmark data, and traceroute information may be used to detect a possible Denial of Service Attack. A detailed log analysis and a passive DNS compromise system may be used to detect a possible unauthorized usage. Finally, a fingerprint (a hash of device configuration data) may be used as a private key to detect a possible spoofing attack. The present invention may be used to help mitigate intrusions and vulnerabilities in IP networks.

Embodiments of the present invention provide a solution for generating interactive cost saving reports on card programs offered to customers. In one embodiment, a method comprises receiving, at a computing device (e.g., a mobile device), characterization data associated with an entity, determining a benchmark data source corresponding to the characterization data, and receiving benchmark data from the determined benchmark data source. The method continues by calculating financial benefit values from the benchmark data and characterization data, generating an interactive savings report, wherein the interactive savings report includes summarized data of the calculated financial benefit values, and displaying the interactive savings report including a plurality of visual metrics corresponding to the summarized data of the calculated financial benefit values, wherein the plurality of visual metrics are interactive.

The invention discloses an image subtitle generation method based on a multi-attention generative adversarial network, belongs to the technical field of image caption generation, and solves the problem that features extracted in the image caption generation method based on the generative adversarial network only contain local points and global information is not captured. A multi-attention mechanism based on local and global information is put forward for the first time to be used for image subtitle generation, and on the basis, a multi-attention generation confrontation image subtitle generation network is put forward and comprises a multi-attention generator and a discriminator. The multi-attention generator is used for generating more accurate sentences, and the multi-attention discriminator is used for judging whether the generated sentences are manually described or generated by a machine. According to the invention, a large number of experimental verifications are carried out onthe proposed framework on the basis of the MSCOCO reference data set, and a very competitive evaluation result is obtained through the evaluation of the MSCOCO subtitle challenge evaluation server.

Apparatuses and methods are provided that can acquire benchmarking data regarding a particular performance metric, compare a pharmacy's performance with the performance of other pharmacies using the benchmarking data, and recommend action plans aimed at improving the pharmacy's performance. The action plans may be customized for the particular pharmacy, so as to provide the pharmacy with a detailed, pharmacy-specific recommendation for improving performance in one or more targeted areas.

The invention discloses a small sample learning algorithm based on covariance measurement, and belongs to the field of computer vision. The algorithm comprises the steps: (1) introducing an interpolation training mechanism to learn migratable knowledge; (2) designing a local covariance representation, and then embedding the local covariance representation into a deep network to learn and express each concept; (3) constructing a covariance measurement layer to measure the distribution consistency between a query sample and the concepts based on the local covariance representation. According tothe method, a novel and concise end-to-end covariance measurement network CovaMNet is provided, a second-order local covariance representation is designed to replace traditional first-order concept representation, and a new covariance measurement function is provided. A comparison experiment result on a plurality of reference data sets is analyzed to obtain that the CovaMNet framework provided bythe invention shows a competitive effect on a general small sample classification task and a fine-grained small sample classification task.

Disclosed is an automatic registration method of aviation and ground LiDAR data based on building contour constraint. The method includes firstly extracting building contours or aviation contours and ground contours for short from aviation and ground LiDAR data; extracting building angular points or aviation angular points and ground angular points for short from the aviation contours and the ground contours; taking the matching degree between the aviation contours and the ground contours as constraint, calculating initial transfer matrixes between the aviation angular points and the ground angular points, and obtaining initial matching angle dot pairs; finally, calculating correction transfer matrixes between the initial matching angle dot pairs by using an iterative closest point (ICP) algorithm, and converting ground dot cloud data to be matched through the initial transfer matrixes and the correction transfer matrixes successively to achieve automatic high-precision registration of the aviation and ground LiDAR data. The method has great advantages in registration reliability and accuracy with contour lines serving as constraint; and simultaneously, based on LiDAR data to be matched and reference LiDAR data, accuracy registration between the aviation and ground LiDAR data can be achieved without other auxiliary data.

The invention aims at providing a method and device for abnormity detection in data synchronization. Specifically, the method comprises the following steps: determining a target range corresponding to synchronous data to be detected in a backup database in the synchronization process of a main database and the backup database; obtaining benchmark data corresponding to the target range; and detecting whether the synchronous data to be detected is consistent to corresponding data in the main database according to the benchmark data. Compared with the prior art, the method and device for abnormity detection in data synchronization have the advantage of efficiently and correctly realizing the consistency detection of the synchronous data between the main database and the backup database in real time through determining the target range to be detected in the backup database and the benchmark data corresponding to the target range in the synchronization process of the main database and the backup database, and detecting whether the synchronous data to be detected in the backup database is consistent to the corresponding data in the main database according to the benchmark data.

The invention provides a 3D target detection method for point cloud screening based on image semantic features. The method comprises the following steps: segmenting image data by a 2D semantic segmentation method to obtain semantic prediction; projecting the generated semantic prediction to an LIDAR point cloud space through a known projection matrix, thereby enabling each point in the point cloudto obtain the semantic category attribute of the corresponding image position; extracting related points of vehicles, pedestrians and riders from the original point cloud and forming a view cone; taking the view cone as the input of a depth 3D target detector, and designing a loss function conforming to the characteristics of the view cone to perform network training. According to the invention,a 3D target detection algorithm for point cloud screening based on image semantic features is designed, so that the 3D detection time is greatly shortened and the calculation requirements are greatlyreduced. Finally, the performance of the method on a reference data set KITTI of 3D target detection shows that the method has good real-time target detection performance.

A scenario modeling engine that generates an enhanced insurance feature recommendation for a group of insured entities or employees is disclosed. The scenario modeling engine may aggregate demographic data regarding the insured entities, receive a user response indicating at least one insurance goal for the group regarding an insurance feature, generate an insurance feature score or profile according to the insurance goal and the received demographic data; generate a first insurance feature recommendation based on the profile; receive industry benchmark data for the insurance feature and report the first insurance feature recommendation compared with the industry benchmark; receive user input indicating revision for the insurance feature; and report a set of insurance features or an insurance plan based on the revision for the insurance feature. A bid or solicitation for a bid for an insurance policy may be automatically sent accordingly.

The invention belongs to the technical field of image processing, and specifically relates to an improved KCF tracking algorithm for target occlusion. Through the improvement of the KCF tracking algorithm, the anti-occlusion performance of the original algorithm is improved, and the tracking accuracy and the success rate of the algorithm are improved. A target occlusion judging method, a trackingfailure judging method and a template matching target re-positioning algorithm are added to the algorithm to improve the anti-occlusion performance of the original KCF tracking algorithm. The improvedalgorithm is verified by using a Benchmark data set and an own shot car video. It can be seen from the experimental results that the algorithm of the invention is more robust against occlusion videosequences.

The invention discloses a self-learning automatic noise reduction system and method of a kitchen ventilator. The system mainly comprises a flue gas turbine, a microphone and a loudspeaker are arranged on the flue gas turbine, after the flue gas turbine is assembled, and before the turbine leaves the factory, the turbine is subject to primary pre-alignment, and the pre-alignment data serve as following field actual running preset benchmark data, and after the flue gas turbine is mounted at the user home, the flue gas turbine runs at the different rotating speeds; flue gas turbine noise and environment noise at different rotating speed segments are collected through the microphone, collected measured data are fed back to a signal processing module on the glue gas turbine, and a self-learning noise reduction algorithm module is combined with the measured data of the flue gas turbine noise and the environment noise to generate a corresponding sound waveform according to the benchmark data preset in a pre-alignment module; the sound waveform is output through the loudspeaker and is used for offsetting most part of noise. The system has the beneficial effects that the main noise component can be effectively removed, a certain obvious effect is achieved, a corresponding sound field and a complex algorithm do not need to be built, and meanwhile, expensive hardware circuit supporting does not need to be built.

The invention provides a task plan adjustment method of a wearable device, and the related device. The method comprises the steps of receiving user activity data corresponding to task units and obtained and uploaded by the wearable device according to an original task plan table; performing mathematic processing on related user activity data by taking the task units as units to determine executive result data of the task units; evaluating the executive result data corresponding to the task units according to evaluation reference data corresponding to the corresponding task units, and determining evaluation result data corresponding to the evaluation reference data; adjusting settings of the task units corresponding to the evaluation result data in the original task plan table according to preset conditions corresponding to the task units, and generating a new task plan table; and pushing the new task plan table to the wearable device. According to the method and the device, a task plan selected by a user is reasonably changed according to an actual condition, so that the purpose of healthily and effectively developing a habit is achieved.