61results about How to "Avoid prediction" patented technology

An automated customer support tool. A web server receives one or more queries regarding the product from a customer. A categorizer coupled to the web server categorizes the received queries based on a type of query, to determine whether the answers to the queries can be automatically communicated to the customer. An FAQ extractor extracts one or more corresponding product FAQs from respective product FAQ data bases. A key-word extractor extracts one or more key-words from the received queries and the extracted product FAQs. The key-word extractor further transforms the extracted key words to unique numerical representations. An analyzer coupled to the key-word extractor represents the transformed key-words into respective query and product FAQ vector forms. The analyzer further applies a convolution algorithm to each of the query vector forms, with each of the product FAQ vector forms separately and obtains one or more appropriate answers to the queries. Then the analyzer automatically communicates the answers to the queries received from the customer.

The invention discloses a communication equipment operation monitoring and predicting system based on big data, relates to the technical field of communication equipment monitoring and predicting, and solves the technical problem that communication equipment cannot be predicted synchronously when operation monitoring is carried out on communication equipment in the prior art. Parameters, needing to be maintained, of the operation monitoring abnormal equipment are obtained through data analysis, the maintenance efficiency and accuracy of the abnormal equipment are improved, meanwhile, performance detection is conducted on the operation monitoring normal equipment through data analysis, and whether the current operation monitoring normal equipment has a fault risk or not is judged; and performing performance prediction on the corresponding performance prediction equipment, and judging the real-time operation efficiency of the normal operation monitoring equipment, thereby performing performance prediction on the normal operation monitoring equipment, timely and accurately analyzing the operation state of the equipment, reducing the influence on a communication network when the equipment fails, and enhancing the stability of the communication network.

The invention discloses a relation visual attention mechanism-based scene graph generation method and mainly aims to solve the problem that redundant relation prediction and interpretability are poorin the prior art. According to the embodiments of the invention, the method includes the following steps of: 1) obtaining the category and boundary frame of targets in images through target detection,and establishing a full-connection relation graph; 2) carrying out sparsification on the relational graph through analyzing a data set to obtain sparse relational graph representations; (3) learninga relation attention transfer function through alternate iteration, transferring subjects and objects to relation occurrence sites through union set features, and learning accurate relation representations; and 4) classifying the learned relation representations, and combining the relation representations into a final scene graph. According to the method, on the basis of the internal relationshipof the relation occurrence of two targets, a relation attention mechanism is established to accurately pay attention to a region where relations occur; a scene graph is accurately generated; the interpretability of a network is improved; and the method can be used for image description and visual question and answer tasks.

A method includes, in a processor, processing a sequence of pre-compiled instructions by an instruction pipeline of the processor. A first block of instructions is identified in the instructions flowing via the pipeline. The first block includes a conditional branch instruction that conditionally diverges execution of the instructions into at least first and second flow-control traces that differ from one another in multiple instructions and converge at a given instruction that is again common to the first and second flow-control traces. A second block of instructions, which is logically equivalent to the first block but replaces the first and second flow-control traces by a single flow-control trace, is created by the processor at runtime. The pipeline is caused to execute the second block instead of the first block.

The invention discloses a bus transfer method based on a bidirectional search and estimation function pruning mode. The bus transfer method includes the following steps: 1) according to a starting station and a terminal station transmitted from a client, a server performs calculation of a straightforward line and a special bus line, calculation of a one-time transfer line, calculation of one-timetransfer line having a segment of walking before transfer, calculation of two-time of transfer line, calculation of two-time of transfer line having a segment of walking before the first time of transfer, calculation of two-time of transfer line having a segment of walking before the second time of transfer, and calculation of two-time of transfer line having a segment of walking both before two times of transfer according to the following methods respectively; and 2) the server calculates the required time of the bus taking scheme available in the step 1 according to the starting station andthe terminal station transmitted from the client, and recommends the bus taking scheme with the least time consumption to a client. The bus transfer method based on a bidirectional search and estimation function pruning mode can improve the going-out efficiency of citizens.

Presented herein are systems, methods, and apparatus for improving performance of video decoders during rewind and fast forward operation. Video decoder performance is improved by avoiding repetitive decoding of prediction pictures. When a decoded prediction picture is stored in a frame buffer, techniques are presented for decoding multiple pictures in the rewind order which are dependent thereon, displaying the picture directly from the frame buffer, and setting one type of prediction picture as another type of prediction picture.

A rubber dam (1) used for protecting a patient (2) during dental procedures. The dam (1) includes a symmetrical round frame (3) which supports a flexible membrane (11). A prepunched hole (19) located approximately 0.5 inch from the membrane center (18) permits rotation of the frame (3) in order to position the hole (19) over the desired tooth (13) without causing the relative position of the frame (3) with respect to the patient's face to be substantially altered.

The invention provides a blood vessel segmentation method and device. The method comprises the following steps: inputting a CTA image into a first convolutional network, and performing blood vessel pre-segmentation based on multi-scale feature extraction; constructing a blood vessel distribution diagram; inputting the CTA image into a second convolutional network, multiplying the output of the second convolutional network by the input CTA image, and inputting the multiplied result into a third convolutional network; enabling the third convolutional network and a graph convolutional network to perform feature interaction based on bidirectional mapping, perform global feature modeling based on a blood vessel distribution graph to capture multi-scale local and global features, and realize blood vessel region prediction; and fusing the multi-scale features to predict a blood vessel segmentation result. According to the invention, blood vessel segmentation is carried out based on cross-network multi-scale feature fusion and local feature and global feature fusion, and compared with blood vessel segmentation based on local features in the prior art, the method can effectively prevent the model from predicting a segmentation region which does not conform to a blood vessel structure, thereby improving the accuracy of blood vessel segmentation.

A method for coding a current block is disclosed. The method comprises:determining a prediction block from the current block,determining a residue block by extracting from the current block the prediction block, andcoding the residue block.The prediction block from the current block is determined according to the following steps for:determining an initial prediction block from motion data and at least one reference image previously coded and reconstructed,applying an atomic decomposition method on a vector of data Ycp, comprising the image data of neighbouring blocks of the current block previously coded and reconstructed and the data of the initial prediction block, andextracting from the decomposed vector the data corresponding to the current block, the extracted data forming the prediction block.

Dual-mode compression of images / videos for reliable real-time transmission includes a scalable block-based video compression. The method provides graceful quality degradations in case of channel bandwidth changes. Prediction for encoding is avoided, and the process is less complex with less memory requirements. The method involves processing each block independently with the advantage that errors are not able to propagate beyond current block boundaries. Two different encoding modes are used, natural image coding for acquired images and graphics coding for computer generated images. The proper encoding mode is chosen by comparing performance of the two modes for a particular block to be encoded. An embedded bitstream is generated by the encoder in accordance to the available channel bandwidth.

A picture encoding and decoding method, a picture encoding and decoding device and a network system are provided, in which, the picture encoding method includes: determining a prediction block used by a picture block according to a division manner of the picture block; determining a corresponding division level in the picture block or the prediction block of transform blocks corresponding to the picture block or the prediction block, in which, the transform blocks corresponding to the picture block or the prediction block include one or more transform blocks; determining a size of the transform blocks corresponding to the picture block or the prediction block according to the division manner of the picture block and the division level; determining identification information for identifying the division level corresponding to the transform blocks corresponding to the picture block or the prediction block; and writing the identification information into a code stream.