Abstract: A highlight learning technique is provided to detect and identify highlights in sports videos. A set of event models are calculated from low-level frame information of the sports videos to identify recurring events within the videos. The event models are used to characterize videos by detecting events within the videos and using the detected events to generate an event vector. The event vector is used to train a classifier to identify the videos as highlight or non-highlight.

Abstract: Methods, systems, and media for determining and presenting information related to embedded sound recordings are provided.

Abstract: Methods, systems, and media for determining and presenting information related to embedded sound recordings are provided.

Abstract: Methods, systems, and media for determining and presenting information related to embedded sound recordings are provided.

Abstract: A highlight learning technique is provided to detect and identify highlights in sports videos. A set of event models are calculated from low-level frame information of the sports videos to identify recurring events within the videos. The event models are used to characterize videos by detecting events within the videos and using the detected events to generate an event vector. The event vector is used to train a classifier to identify the videos as highlight or non-highlight.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for associating still images and videos. One method includes receiving a plurality of images and a plurality of videos and determining whether the images are related to the videos. The determining includes, for an image and a video, extracting features from the image and extracting features frames of the video, and comparing the features to determine whether the image is related to the video. The method further includes maintaining a data store storing data associating each image with each video determined to be related to the image.

Abstract: A highlight learning technique is provided to detect and identify highlights in sports videos. A set of event models are calculated from low-level frame information of the sports videos to identify recurring events within the videos. The event models are used to characterize videos by detecting events within the videos and using the detected events to generate an event vector. The event vector is used to train a classifier to identify the videos as highlight or non-highlight.

Abstract: Methods, systems, and media for determining and presenting information related to embedded sound recordings are provided.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for associating still images and videos. One method includes receiving a plurality of images and a plurality of videos and determining whether the images are related to the videos. The determining includes, for an image and a video, extracting features from the image and extracting features frames of the video, and comparing the features to determine whether the image is related to the video. The method further includes maintaining a data store storing data associating each image with each video determined to be related to the image.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for associating still images and videos. One method includes receiving a plurality of images and a plurality of videos and determining whether the images are related to the videos. The determining includes, for an image and a video, extracting features from the image and extracting features frames of the video, and comparing the features to determine whether the image is related to the video. The method further includes maintaining a data store storing data associating each image with each video determined to be related to the image.

Abstract: In one example, a method includes receiving, by a computing device, an indication of a detected force applied to the computing device. The method further comprises determining, by the computing device, that the detected force matches a corresponding input that the computing device associates with a corresponding function that is executable by the computing device. The method further comprises generating, by the computing device and in response to determining that the detected force matches the corresponding input and, a non-visual output based on the corresponding function.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for associating still images and videos. One method includes receiving a plurality of images and a plurality of videos and determining whether the images are related to the videos. The determining includes, for an image and a video, extracting features from the image and extracting features frames of the video, and comparing the features to determine whether the image is related to the video. The method further includes maintaining a data store storing data associating each image with each video determined to be related to the image.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for associating still images and videos. One method includes receiving a plurality of images and a plurality of videos and determining whether the images are related to the videos. The determining includes, for an image and a video, extracting features from the image and extracting features frames of the video, and comparing the features to determine whether the image is related to the video. The method further includes maintaining a data store storing data associating each image with each video determined to be related to the image.

Abstract: The disclosure includes a system and method for recommending multi-party communication sessions. The system includes: a user model module, a content model module and a determination module. The user model module determines a user social interest model based at least in part on user profile data. The content data module determines interest data describing content data associated with one or more content identifiers. The content model module determines one or more content models describing the content data. The determination module determines whether a match exists between the user social interest model and the one or more content models. The determination module determines the one or more multi-party communication sessions for the recommendation responsive to determining the match exists between the user social interest model and the one or more content models.

Abstract: User engagement in unwatched videos is predicted by collecting and aggregating data describing user engagement with watched videos. The data are normalized to reduce the influence of factors other than the content of the videos on user engagement. Engagement metrics are calculated for segments of watched videos that indicate user engagement with each segment relative to overall user engagement with the watched videos. Features of the watched videos within time windows are characterized, and a function is learned that relates the features of the videos within the time windows to the engagement metrics for the time windows. The features of a time window of an unwatched video are characterized, and the learned function is applied to the features to predict user engagement to the time window of the unwatched video. The unwatched video can be enhanced based on the predicted user engagement.

Abstract: A highlight learning technique is provided to detect and identify highlights in sports videos. A set of event models are calculated from low-level frame information of the sports videos to identify recurring events within the videos. The event models are used to characterize videos by detecting events within the videos and using the detected events to generate an event vector. The event vector is used to train a classifier to identify the videos as highlight or non-highlight.

Abstract: This disclosure relates to allowing retailers of all sizes to log information regarding shoppers and to provide shoppers with product recommendations. The product recommendations may be responsive to logged shopper information, such as shoppers' current or previous browsing habits, for example. The product recommendations also may be responsive to other sources of information, such as search trends on search engines and/or information provided in blogs, social network websites, and/or other websites. A service provider, which is not a retailer, may provide the product recommendations for the retailer, either directly or by providing a script for embedding within the retailer's website. For example, the retail website may include a logging script and/or a product recommendation script, which communicate with one or more backend computing devices to provide the product recommendations.

Abstract: A highlight learning technique is provided to detect and identify highlights in sports videos. A set of event models are calculated from low-level frame information of the sports videos to identify recurring events within the videos. The event models are used to characterize videos by detecting events within the videos and using the detected events to generate an event vector. The event vector is used to train a classifier to identify the videos as highlight or non-highlight.

Abstract: User engagement in unwatched videos is predicted by collecting and aggregating data describing user engagement with watched videos. The data are normalized to reduce the influence of factors other than the content of the videos on user engagement. Engagement metrics are calculated for segments of watched videos that indicate user engagement with each segment relative to overall user engagement with the watched videos. Features of the watched videos within time windows are characterized, and a function is learned that relates the features of the videos within the time windows to the engagement metrics for the time windows. The features of a time window of an unwatched video are characterized, and the learned function is applied to the features to predict user engagement to the time window of the unwatched video. The unwatched video can be enhanced based on the predicted user engagement.

Abstract: This technology may generate recommendations of relevant content, based on determining intersections among one or more user interest profiles of users, who interact either synchronously or asynchronously. This technology may retrieve interest profiles for particular users, determine intersections among all user interest profiles (or among individual content recommendations), and create a group interest profile, update the particular users' interest profiles based on the group interest profile created, and generate recommendations of content that is determined to be relevant based on the group interest profile. A user may select items from these recommendations of content, add the user-selected content recommendations to a common group pool, generate a group interest profile based on the common group pool, and generate recommendations of content based on the group interest profile. Scores for the recommendations of content may be calculated and the top scoring ones may be displayed to the users in the group.