Abstract: A bot traffic detection system detects scripted network traffic. The bot traffic detection system may use a one-sided unsupervised machine learning technique to estimate distributions for human, non-scripted traffic (clean distributions). The clean distributions may be dynamically updated based on the latest traffic patterns. To estimate the clean distributions the bot traffic detection system may identify, for a certain subset of network traffic, feature values of the certain subset of network traffic that do not include bot traffic (clean buckets). Using clean traffic may provide more robust and stable behavior that can be tracked over time. Using the clean distributions, the bot traffic detection system may generate a rules table that indicates a likelihood that network traffic with a given combination of feature values is scripted network traffic. The bot traffic detection system may apply the rules table in real time to identify scripted network traffic.

Abstract: A bot traffic detection system detects scripted network traffic. The bot traffic detection system may use a one-sided unsupervised machine learning technique to estimate distributions for human, non-scripted traffic (clean distributions). The clean distributions may be dynamically updated based on the latest traffic patterns. To estimate the clean distributions the bot traffic detection system may identify, for a certain subset of network traffic, feature values of the certain subset of network traffic that do not include bot traffic (clean buckets). Using clean traffic may provide more robust and stable behavior that can be tracked over time. Using the clean distributions, the bot traffic detection system may generate a rules table that indicates a likelihood that network traffic with a given combination of feature values is scripted network traffic. The bot traffic detection system may apply the rules table in real time to identify scripted network traffic.

Abstract: Technologies pertaining to analyzing content extracted from web pages by a static crawler to determine whether respective web pages are members of a malware distribution network (MDN) are described. A set of features is learned based upon output of a dynamic crawler over known landing pages of a particular MDN, wherein the set of features are indicative of membership in the MDN. Using such set of features, additional members of the MDN (not subjected to crawling by a dynamic crawler) are identified.

Abstract: Technologies pertaining to analyzing content extracted from web pages by a static crawler to determine whether respective web pages are members of a malware distribution network (MDN) are described. A set of features is learned based upon output of a dynamic crawler over known landing pages of a particular MDN, wherein the set of features are indicative of membership in the MDN. Using such set of features, additional members of the MDN (not subjected to crawling by a dynamic crawler) are identified.

Abstract: Technologies pertaining to analyzing content extracted from web pages by a static crawler to determine whether respective web pages are members of a malware distribution network (MDN) are described. A set of features is learned based upon output of a dynamic crawler over known landing pages of a particular MDN, wherein the set of features are indicative of membership in the MDN. Using such set of features, additional members of the MDN (not subjected to crawling by a dynamic crawler) are identified.

Abstract: A current prefix character string representing a prefix of a proposed password may be obtained from a user input device. A prediction of a most likely next character of the proposed password may be determined, based on applying a set of heuristics to the current prefix character string. A response indicating an impact on a security strength of the proposed password may be determined, based on a selection of the predicted most likely next character.

Abstract: Technologies pertaining to analyzing content extracted from web pages by a static crawler to determine whether respective web pages are members of a malware distribution network (MDN) are described. A set of features is learned based upon output of a dynamic crawler over known landing pages of a particular MDN, wherein the set of features are indicative of membership in the MDN. Using such set of features, additional members of the MDN (not subjected to crawling by a dynamic crawler) are identified.

Abstract: A current prefix character string representing a prefix of a proposed password may be obtained from a user input device. A prediction of a most likely next character of the proposed password may be determined, based on applying a set of heuristics to the current prefix character string. A response indicating an impact on a security strength of the proposed password may be determined, based on a selection of the predicted most likely next character.

Abstract: A “media stream customizer” customizes buffered media streams by inserting one or more media objects into the stream to maintain an approximate buffer level. Specifically, when media objects such as songs, jingles, advertisements, etc., are deleted from the buffered stream (based on some user specified preferences), the buffer level will decrease. Therefore, over time, as more objects are deleted, the amount of the media stream being buffered continues to decrease, thereby limiting the ability to perform additional deletions from the stream. To address this limitation, the media stream customizer automatically chooses one or more media objects to insert back into the stream, and ensures that the inserted objects are consistent with any surrounding content of the media stream, thereby maintaining an approximate buffer level. In addition, the buffered content can also be stretched using pitch preserving audio stretching techniques to further compensate for deletions from the buffered stream.

Abstract: A “Music Mapper” automatically constructs a set coordinate vectors for use in inferring similarity between various pieces of music. In particular, given a music similarity graph expressed as links between various artists, albums, songs, etc., the Music Mapper applies a recursive embedding process to embed each of the graphs music entries into a multi-dimensional space. This recursive embedding process also embeds new music items added to the music similarity graph without reembedding existing entries so long a convergent embedding solution is achieved. Given this embedding, coordinate vectors are then computed for each of the embedded musical items. The similarity between any two musical items is then determined as either a function of the distance between the two corresponding vectors. In various embodiments, this similarity is then used in constructing music playlists given one or more random or user selected seed songs or in a statistical music clustering process.

Abstract: A system for distributing digital content over a computer network (e.g., the Internet) uses certificates to establish a trust relationship between a content provider and a display device. The certificates identify the display device and the content provider as well as unique characteristics of the distribution. For example, the content provider may be a book publisher and the display device may be a printer/binder.

Abstract: Many media streams contain “objects” that repeat. Repeating objects in a media stream are defined as any section of non-negligible duration, i.e., a song, video, advertisement, jingle, etc., which would be considered to be a logical unit by a human listener or viewer. An “object controller” identifies such repeating objects as they occur, and provides an interactive user interface for allowing users to specify how individual repeating objects are to be handled either in real time, or upon subsequent occurrences of particular repeating objects. In general, the object controller includes a mechanism for identifying repeating objects, a mechanism for identifying temporal endpoints of those objects, a user interface for specifying actions to be taken when a particular object repeats within a media stream, and, in one embodiment, a buffer having sufficient length to allow for real-time deletion of objects from the media stream without obvious interruption in the str.

Abstract: The present invention provides a large capacity, user defined audio content delivery system. The system delivers uninterrupted music and delivers information content (e.g., news by evaluating and encoding an input audio stream while outputting another stream. Undesirable audio content (e.g., advertisements and unwanted news) are not present in the output audio stream as only desired portions of information content are stored for playback on demand. The invention also includes a user interface that is simple enough to facilitate utilization of the audio system in an automobile and employs standard hardware available in typical computing and/or personal digital assistant equipment. Additionally, the audio system can be portable (e.g., as portable as a personal digital assistant) and can be updated in real time or off line via a personal computer.

Abstract: A “media stream customizer” customizes buffered media streams by inserting one or more media objects into the stream to maintain an approximate buffer level. Specifically, when media objects such as songs, jingles, advertisements, etc., are deleted from the buffered stream (based on some user specified preferences), the buffer level will decrease. Therefore, over time, as more objects are deleted, the amount of the media stream being buffered continues to decrease, thereby limiting the ability to perform additional deletions from the stream. To address this limitation, the media stream customizer automatically chooses one or more media objects to insert back into the stream, and ensures that the inserted objects are consistent with any surrounding content of the media stream, thereby maintaining an approximate buffer level. In addition, the buffered content can also be stretched using pitch preserving audio stretching techniques to further compensate for deletions from the buffered stream.

Abstract: Many media streams contain “objects” that repeat. Repeating objects in a media stream are defined as any section of non-negligible duration, i.e., a song, video, advertisement, jingle, etc., which would be considered to be a logical unit by a human listener or viewer. An “object controller” identifies such repeating objects as they occur, and provides an interactive user interface for allowing users to specify how individual repeating objects are to be handled either in real time, or upon subsequent occurrences of particular repeating objects. In general, the object controller includes a mechanism for identifying repeating objects, a mechanism for identifying temporal endpoints of those objects, a user interface for specifying actions to be taken when a particular object repeats within a media stream, and, in one embodiment, a buffer having sufficient length to allow for real-time deletion of objects from the media stream without obvious interruption in the stream.

Abstract: System and method for electronically transferring data through a communications connection in a transparent manner such that the data transfer does not interfere with other traffic sharing the connection. The invention transfers data using bandwidth of the connection that other traffic are not using. If other traffic desires to use the bandwidth currently being used by the invention, the invention relinquishes the bandwidth to the other traffic and retreats to avoid bandwidth contention. Although a retreat may cause gaps in the data transferred, a key aspect of the invention is that any missing data due to these gaps is recovered easily and in a bandwidth-efficient way using novel error correction and recovery.

Abstract: An “object extractor” automatically identifies and segments repeating media objects in a media stream. “Objects” are any section of non-negligible duration, i.e., a song, video, advertisement, jingle, etc., which would be considered to be a logical unit by a human listener or viewer. Identification and segmentation of repeating objects is achieved by directly comparing sections of the media stream to identify matching portions of the stream, then aligning the matching portions to identify object endpoints. Alternately, a suite of object dependent algorithms is employed to target particular aspects of the stream for identifying possible objects within the stream. Confirmation of possible objects as repeating objects is achieved by automatically searching for potentially matching objects in a dynamic object database, followed by a detailed comparison to one or more of the potentially matching objects. Object endpoints are then determined by automatic alignment and comparison to other copies of that object.

Abstract: A “composite signal generator” automatically combines two or more copies of a signal to produce a composite that is better than the individual copies. For example, given two or more input images of a scene captured from approximately the same viewpoint, the composite signal generator automatically produces a composite image having reduced or eliminated areas of occlusion with respect to any occlusions existing in the input images. First, the input images are registered using conventional image registration techniques. Differences between the registered images are then used to identify regions of potential occlusion in one or more of the images. A determination of which image is actually occluded is made by identifying which image has a larger discontinuity along a border of the potentially occluded regions. A composite image is then created by choosing one image and mosaicing parts from the other images where it is occluded and they are not.

Abstract: The “repeating object extractor” described herein automatically segments and identifies audio objects such as songs, station identifiers, jingles, advertisements, emergency broadcast signals, etc., in an audio stream. The repeating object extractor performs a computationally efficient joint segmentation and identification of the stream even in an environment where endpoints of audio objects are unknown or corrupted by voiceovers, cross-fading, or other noise. Parametric information is computed for a portion of the audio stream, followed by an initial comparison pass against a database of known audio objects to identify known audio objects which represent potential matches to the portion of the audio stream. A detailed comparison between the potential matches and the portion of the audio stream is then performed to confirm actual matches for segmenting and identifying the audio objects within the stream, an alignment between matches is then used to determine extents of audio objects in the audio stream.

Abstract: A background color estimation system and method for estimating a background color of a scanned image. The background color estimation method includes generating a list of candidate background colors present in the image, determining the most common candidate color from the list, and designating the most common candidate background color as the estimated background color. Candidate colors are extracted from the scanned image by examining lines of pixels in the image and generating frequency distributions (such as histograms) of pixel colors within the lines. The frequency value of the most common pixel color in the frequency distribution is compared to a frequency threshold. If the frequency value is greater than the frequency threshold, the most common pixel color is considered a candidate background color. In addition, a variance of the estimated background color is obtained and can be used to intelligently select a threshold for additional image processing techniques.