Abstract: Some embodiments of the invention provide techniques for an advertisement mask and a target media content being jointly encoded, transformed, and progressively rendered for presentation to a user. Specifically, a request for the target media content by a user is received. The target media content and the advertisement mask are scaled and divided into equally sized blocks and, further, jointly encoded into a compressed media file. Transformation of the content of the compressed media file is followed by progressively loading and rendering the advertisement mask and a partially obscured view of the media content. As the target media content is transmitted and fully rendered, the advertisement mask gradually decreases in opacity until it is removed for presentation of the fully-rendered target media content.

Abstract: Software for supervised learning extracts a set of pixel-level features from each source image in collection of source images. Each of the source images is associated with a thumbnail created by an editor. The software also generates a collection of unique bounding boxes for each source image. And the software calculates a set of region-level features for each bounding box. Each region-level feature results from the aggregation of pixel values for one of the pixel-level features. The software learns a regression model, using the calculated region-level features and the thumbnail associated with the source image. Then the software chooses a thumbnail from a collection of unique bounding boxes in a new image, based on application of the regression model.

Abstract: Techniques are provided for efficiently identifying relevant product images based on product items detected in a query image. In general, a query image may represent a digital image in any format that depicts a human body and one or more product items. For example, a query image may be an image for display on a webpage, an image captured by a user using a camera device, or an image that is part of a media content item, such as a frame from a video. Product items may be detected in a query image by segmenting the query image into a plurality of image segments and clustering one or more of the plurality image segments into one or more image segment clusters. The resulting image segments and image segment clusters may be used to search for visually similar product images.

Abstract: Systems and methods are described for determining manipulation history among a plurality of images. The described techniques include selecting a pair of images from the plurality of images, detecting one or more manipulations operable to transform one of the images to the other, and based on the manipulations detected, determining a parent-child relationship between the pair or pairs of images. The described techniques can further include repeating the selecting two images, detecting manipulations, and determining the parent-child relationship for each pairs of images in the plurality of images, constructing a visual migration map for the images, and presenting the visual migration map in a user readable format.

Abstract: Software for supervised learning extracts a set of pixel-level features from each source image in collection of source images. Each of the source images is associated with a thumbnail created by an editor. The software also generates a collection of unique bounding boxes for each source image. And the software calculates a set of region-level features for each bounding box. Each region-level feature results from the aggregation of pixel values for one of the pixel-level features. The software learns a regression model, using the calculated region-level features and the thumbnail associated with the source image. Then the software chooses a thumbnail from a collection of unique bounding boxes in a new image, based on application of the regression model.

Abstract: Disclosed herein are aspects associated with contextual, or related, media enrichment presentation item of a media object served via the internet. A request to annotate a media object in connection with the media object's presentation is received, and a media object identifier and a profile identifier are obtained. The media object's information is retrieved using the media object identifier, and a profile is retrieved using the profile identifier. A response including one or more references to one or more media enrichment presentation items is transmitted, each reference to a media enrichment presentation item comprising information for use in retrieving the media enrichment presentation item for presentation in connection with presentation of the media object.

Abstract: A system and method for using topic messages to understand media relating to an event. An event (such as a live event or broadcast event), is selected where at least a portion of the content of the event is stored as at least one media object, such as, for example a video or audio recording in digital format. A plurality of topic messages relating to the broadcast event is retrieved over a network and are analyzed to identify a plurality of content segments within the event. The topic messages are further analyzed to identify at least one key actor related to the event. Segment and key actor data relating to the event is stored to a computer readable medium in association with the media objects, whereby the segment data identifies the plurality of content segments within the media objects and the key actor data identifies key actors in the event.

Abstract: Embodiments are directed towards identifying authoritative sources of multimedia content useable in rank ordering class-dependent search-query results of multimedia content. In one embodiment, a citation model is employed for measuring or otherwise determining a strength of an authority to a content source. In one embodiment, a directed graph is constructed over a network of sources based on a propensity of one source to “cite” content provided by another source. In one embodiment, a random walk may be conducted across the network of sources to arrive at authority scores for each source in the network. In another embodiment, a machine-learning algorithm may be used to arrive at authority scores. The authority scores may then be applied for ranking, for example, search-query results, and/or retrieval purposes.

Abstract: Embodiments are directed towards identifying authoritative sources of multimedia content useable in rank ordering class-dependent search-query results of multimedia content. In one embodiment, a citation model is employed for measuring or otherwise determining a strength of an authority to a content source. In one embodiment, a directed graph is constructed over a network of sources based on a propensity of one source to “cite” content provided by another source. In one embodiment, a random walk may be conducted across the network of sources to arrive at authority scores for each source in the network. In another embodiment, a machine-learning algorithm may be used to arrive at authority scores. The authority scores may then be applied for ranking, for example, search-query results, and/or retrieval purposes.

Abstract: This patent discloses a system to compile a landmark image search result. The system may determine a rank of each image within a visual cluster according to at least one of a low-level self-similarity score, a low-level discriminative modeling score, and a point wise linking score. The landmark image search result may be compiled as a function of the rank of each image.

Abstract: This patent discloses a system to compile a landmark image search result. The system may determine a rank of each image within a visual cluster according to at least one of a low-level self-similarity score, a low-level discriminative modeling score, and a point wise linking score. The landmark image search result may be compiled as a function of the rank of each image.

Abstract: The present disclosure is descriptive of discovering structure, content, and context of a media event, e.g., a live media event, using real-time discussions that unfold through short messaging services. Generally, a sampling of short messages of a plurality of users is obtained. The sampling of short messages corresponds to a media event. A segment in the media event is identified using the sampling of short messages, and at least one term taken from the sampling of short messages is identified. The at least one term is indicative of a context of the identified segment.

Abstract: A system and method for using topic messages to understand media relating to an event. An event (such as a live event or broadcast event), is selected where at least a portion of the content of the event is stored as at least one media object, such as, for example a video or audio recording in digital format. A plurality of topic messages relating to the broadcast event is retrieved over a network and are analyzed to identify a plurality of content segments within the event. The topic messages are further analyzed to identify at least one key actor related to the event. Segment and key actor data relating to the event is stored to a computer readable medium in association with the media objects, whereby the segment data identifies the plurality of content segments within the media objects and the key actor data identifies key actors in the event.

Abstract: Systems and methods are described for determining manipulation history among a plurality of images. The described techniques include selecting a pair of images from the plurality of images, detecting one or more manipulations operable to transform one of the images to the other, and based on the manipulations detected, determining a parent-child relationship between the pair or pairs of images. The described techniques can further include repeating the selecting two images, detecting manipulations, and determining the parent-child relationship for each pairs of images in the plurality of images, constructing a visual migration map for the images, and presenting the visual migration map in a user readable format.

Abstract: This patent discloses a system to compile a landmark image search result. The system may determine a rank of each image within a visual cluster according to at least one of a low-level self-similarity score, a low-level discriminative modeling score, and a point wise linking score. The landmark image search result may be compiled as a function of the rank of each image.