Abstract: The present invention performs a method for matching slides to video comprising, generating keyframes from the video, extracting geometric keyframe features from the keyframes and geometric slide features from the slides, and matching the geometric slide features and the geometric keyframe features.

Abstract: Digital data is preserved by archiving on a removable medium. In the long term, the save data bit stream must be correctly interpreted. For a computer program or system to be archived, the bit stream constituting the program must be archived and the code must be executable at restore time. The program that restores the data does not “see” the contents of the data itself, but accesses it by issuing a function call to an executor. A description of which methods are available to restore the information hidden in the data is always available. A text tells the client which functions are available and what their purposes are. The archiving method is based on using a virtual computer instruction set and saving the algorithm as a program written int hat virtual machine language. For machine instructions to be executed many years later, for example 100 years, an emulator of the original machine would be written on the future hardware.

Abstract: A method and structure for processing packets of information over a data processing system is disclosed. The packets have header information indicating a source and a destination of the packets. The method includes receiving a packet from a client, forwarding the packet to a hub of a local area network, changing the header of the packet such that the packet is directed to a server instead of an original destination, processing the packet through the server to produce a processed packet, and forwarding the processed packet to the original destination.

Abstract: A method and structure for identifying database association rules includes mining first database association rules, the first database association rules having ratings with respect to a plurality of metrics and population constraints, selecting second database association rules from the first database association rules, each of the second database association rules having a highest rating with respect to a different metric of the metrics, and interactively changing the metrics and repeating the selecting to identify most important ones of the databases association rules for a given set of metrics.

Abstract: A method and structure for identifying database association rules includes mining first database association rules, the first database association rules having ratings with respect to a plurality of metrics, selecting second database association rules from the first database association rules, each of the second database association rules having a highest rating with respect to a different metric of the metrics, and interactively changing the metrics and repeating the selecting to identify most important ones of the databases association rules for a given set of metrics.

Abstract: A portable computer includes a battery and a connection to an external power source, a two-speed data storage device being supplied power from one or more of the battery and the external power source, and a controller attached to the storage device. With the invention, when the storage device is powered by the internal battery, the controller not only reduces the rotation speed and the clock rate of the storage device, but also reduces the power consumption of the read/write electronics module inside the disk drive by lowering the power supply voltage for the write driver inside said module and lowering the tail currents for the amplifier stages in the readback amplifier inside said module.

Abstract: A magnetic recording channel front-end for a magnetic storage system includes a magnetoresistive element, an interconnect and a readback amplifier. The interconnect, which has a characteristic impedance, couples the magnetoresistive element to the readback amplifier. The readback amplifier includes a gain stage and an active termination. The gain stage, which has an associated impedance, has an input that is connected to the interconnect. The active termination is coupled to the input of the gain stage, such that an input impedance of the readback amplifier, which is formed by a combination of the impedance associated with the gain stage and the active termination, is substantially equal to the characteristic impedance of the interconnect. Additionally, the gain stage generates a first noise signal that has a first magnitude appearing at an output of the gain stage. The active termination generates a second noise signal that has a second magnitude that also appears at the output of the gain stage.

Abstract: A method for forming a termination for a magnetic write head in which a magnetic write coil having an inner turn and an outer turn is formed on a wafer. A bottom capacitor plate for a capacitor on the wafer. A resistor having a first terminal and a second terminal is also formed on the wafer. A dielectric layer is formed over the bottom capacitor plate for the capacitor. A top capacitor plate is formed on the dielectric layer. An overpass lead is formed between the inner turn of the coil and one of the bottom capacitor plate and the top capacitor plate. The first terminal of resistor is connected to the capacitor plate that is not connected to the inner turn of the write coil. The second terminal of the resistor is connected to the outer turn of the write coil. An impedance formed by the write coil, the capacitor and the resistor substantially equals a characteristic impedance Z0 of an interconnect circuit that will be connected to the write coil.

Abstract: Images in an image database are indexed according to semantic labels from a surface color code book in order to allow querying and retrieving of images based on color contend. Regions of unsegmented images in a database are localized and described with semantic labels from a surface color code book. A database index is then built using this image region information and corresponding labels. This pre-generated color-region index allows querying by semantic labels in order to locate and retrieve similar images from a database. Querying for similar images by presenting a target image is also possible using the described color-region index. In this second type of query mode a target image is analyzed to determine semantic labels of color surfaces present in that image and these labels, in turn, are used to locate similar entries in the color-region index.

Abstract: Colored surfaces are represented in an illumination-invariant and pose-invariant manner through a color code book. Patches of each color surface are represented, under well-chosen illumination conditions and other imaging variations, as clusters in an optimal color space. For each surface, the color code book entries represent its clusters' means, covariance matrices, eigen values and eigen vectors. In addition, a corresponding semantic label is given to each surface's color code book entry to simplify image processing operations such as region localization and recognition as well as indexing and querying of image databases.

Abstract: A objective function based on a Bayesian statistical estimation framework is used to determine an optimal model selection by choosing both the optimal number of clusters and the optimal feature set. Heuristics can be applied to find the optimal (or at least sub-optimal) of this objective function in terms of the feature sets and the number of clusters, wherein the maximization of the objective function corresponds to the optimal model structure.

Abstract: Recovery of a filesystem directory structure is performed to restore it to any point in time and also to synchronize a database restore and a filesystem restore to bring the two restores to a database consistent state. A database management system (DBMS) manages external files and hierarchical directory structures to enable recovery and reconciliation of the files and filesystems, under DBMS control, after filesystem crashes. First, a database table, which recorded previous directory creations and deletions, is used to rebuild a filesystem's directory structure to any previous database state and then external file link information is used to restore files to that same state.

Abstract: A surface color code book is used to allow region localization and recognition within images without the need for pre-segmentation of the image. Pixels in an image are first mapped into an optimal color space and assigned a closest color label from the surface color code book. Clusters of adjacent pixels having a similar color label are then tentatively grouped to form a region. These tentative regions are then projected into the optimal color space and a descriptor is formulated for each re-projected cluster. Each descriptor is compared against the color codes in the color code book to determine if the initially assigned color label was correct; if correct, then that region receives both a color label from the color code book and a region identifier.

Abstract: A magnetic recording circuit includes a write driver output stage, a magnetic write head and an interconnect circuit. The write driver output stage has an output and an output impedance. The magnetic write head has an input and an input impedance. The interconnect circuit has an input, an output and a selected characteristic impedance Z0. The input of the interconnect circuit is coupled to the output of the write driver output stage, the output of the interconnect circuit is coupled to the input of the write head. The interconnect circuit can be one of, or both, source-terminated so that the output impedance of the write driver output stage is substantially equal to the selected characteristic impedance Z0 of the interconnect circuit, and load-terminated so that the input impedance of the write head is substantially equal to the characteristic impedance Z0 of the interconnect circuit.

Abstract: A system for secure data entry, includes a virtual keypad having a plurality of keys, a mechanism for determining to which virtual key of the keypad a user is looking, and an actuator, operable by the user, for confirming key selection. Another system for secure data entry, includes a virtual keypad having a plurality of keys, an eye tracker for tracking eye movement of a user, and for receiving a coded input from an eye movement of the user gazing upon at least a selected one of the keys of the virtual keypad, and an actuator for being selectively actuated by the user upon confirmation of the coded input by the user.

Abstract: A technique for clustering data points in a data set that is arranged as a matrix having n objects and m attributes. Each categorical attribute of the data set is converted to a 1-of-p representation of the categorical attribute. A converted data set A is formed based on the data set and the 1-of-p representation for each categorical attribute. The converted data set A is compressed using, for example, a Goal Directed Projection compression technique or a Singular Value Decomposition compression technique, to obtain q basis vectors, with q being defined to be at least m+1. The transformed data set is projected onto the q basis vectors to form a data matrix having at least one vector, with each vector having q dimensions. Lastly, a clustering technique is performed on the data matrix having vectors having q dimensions.

Abstract: A system and method for data mining is provided in which temporal patterns of itemsets in transactions having unexpected support values are identified. A surprising temporal pattern is an itemset whose support changes over time. The method may use a minimum description length formulation to discover these surprising temporal patterns.

Abstract: A system and method for determining the significance of association rules which are mined from a dataset is provided. Predictive association rules may also be generated based on the significance of an association rule. The statistical significance of an association rule may be used to estimate the number of false discoveries in a dataset or to rank discovered association rules by statistical significance to permit the user of the system to view the most statistically significant association rules first.

Abstract: A method and apparatus for efficiently matching a large collection of user profiles against a large volume of data in a webcasting system. The invention generally includes in one embodiment four steps to parallelize the profiles. First, an initial profile set is partitioned into several subsets also referred to as sub-partitions using various heuristic methods. Second, each sub-partition is mapped onto one or more independent processing units. Each processing unit is not required to have equal processing performance. However, for best performance results, subset data should be mapped in one embodiment where the subset with a highest cost is mapped to a fastest processor, and the next highest cost subset mapped to the next fastest processor. Where appropriate, the invention evaluates the relative subset processing speed of each processor and adjusts future subset mapping based upon these evaluations.