Abstract: The disclosure includes a system and method for identifying advertising content and providing a user of a mobile device with a channel related to the advertising content. A recognition application receives advertising content comprising a series of frames, identifies images corresponding to the frames within the advertising content, identifies a first set of feature points from a first image of the advertising content, determines whether the first set of feature points matches a second set of feature points stored in a database, and in response to determining that the first set of feature points fails to match the second set of feature points, identifies the first image as being unique, indexes the first image and the first set of feature points and associates the first image with a channel.

Abstract: A crowdsourced question and answer platform enhanced with image recognition technology is disclosed. The platform generates and organizes conversations surrounding an object recognized from an image uploaded by a user. Generating and organizing conversations includes receiving an image a from a user, analyzing the image to recognize an object in the image, receiving a comment related to the recognized object in the image from the user, generating content related to the recognized object in the image based on the received comment and adding the content related to the recognized object in the image to a conversation corresponding to the recognized object in the image.

Abstract: The disclosure includes a system and method for recommending an advertising campaign based on which channels are popular. A recognition application receives captured images from user devices, the captured images including information about products, identify indexed images that match the captured images, identify metadata about the captured images including information associated with products in the captured images, determine popularity of each channel based on a number of interactions that users have had with channels during a time period based on the metadata, receive an advertisement and provide on the one or more channels the advertisement based on the popularity of each channel.

Abstract: A computer-implemented method for monetizing future location of users includes identifying indicator(s) retrieved from log data or an electronic declaration that indicate a possibility that a user will be at a specific location in the future; generating a probability that the user will be at the specific location in the future based on the indicator(s); and charging an advertiser a cost for an impression to which to serve a location-specific advertisement targeted to the user, the cost charged being greater when the probability is greater that the user will be at the specific location in the future. The method may be extended to targeting a group of users that expect to be at the same location at the future point in time. Monetization of future locations may depend on the quality and/or quantity of the log data and declarations of future locations and on a number of users in a group of users if monetized according to the group.

Abstract: The disclosure includes a system and method for grouping users into one or more categories for delivering targeted advertisements. A media application receives a captured image from a user, identifies an indexed image matching the captured image, retrieves metadata describing the captured image responsive to identifying the indexed image, generates a user profile for the user based on the metadata describing the captured image and a history of at least one previously captured image associated with the user, identifies a product in the captured image based on the metadata describing the captured image, provides the user with information about the product, updates the user profile based on how the user interacts with the product, and categorizes the user into one or more groups based on the user profile.

Abstract: The disclosure includes a system and method selecting advertisements for users based on image capturing patterns across different media. A media application receives a captured image including information about a product from a user, identifying an indexed image matching the captured image including information about the product, retrieving metadata describing the captured image including information about the product responsive to identifying the indexed image, retrieving a user profile stored for the user, identifying an advertisement from a set of advertisements based on the user profile and the metadata describing the captured image including information about the product and providing the user with the advertisement based on the user profile.

Abstract: Techniques are described for providing automated recommendations of real-world locations, such as businesses, for users to visit based at least in part on historical location-preference information. The historical location-preference information used by the recommendation system may include the historical location-preference information of the person that requests the recommendation, other people explicitly identified as participants by the requestor, and/or other people implicitly determined to be participants. The historical location-preference information may be explicit, such as “check-ins” or reviews, or implicit. Implicit participants may be identified in a variety of ways, including social network relationships and the context in which the recommendation request is submitted.

Abstract: A crowdsourced question and answer platform enhanced with image recognition technology is disclosed. The platform generates and organizes conversations surrounding an object recognized from an image uploaded by a user. Generating and organizing conversations includes receiving an image a from a user, analyzing the image to recognize an object in the image, receiving a comment related to the recognized object in the image from the user, generating content related to the recognized object in the image based on the received comment and adding the content related to the recognized object in the image to a conversation corresponding to the recognized object in the image.

Abstract: The disclosure includes a system and method for identifying advertising content and providing a user of a mobile device with a channel related to the advertising content. A recognition application receives advertising content comprising a series of frames, identifies images corresponding to the frames within the advertising content, identifies a first set of feature points from a first image of the advertising content, determines whether the first set of feature points matches a second set of feature points stored in a database, and in response to determining that the first set of feature points fails to match the second set of feature points, identifies the first image as being unique, indexes the first image and the first set of feature points and associates the first image with a channel.

Abstract: The disclosure includes a system and method for identifying discussion groups related to a video that a user would find to be interesting. A recognition application receives a captured image from a user, identifies an indexed image that matches the captured image, identifies a timestamp for the indexed image, determines a segment in the video that encompasses the timestamp, determines whether the user is interested in the segment of the video based on the timestamp and a user profile and provides suggested discussion groups to the user that are ranked based on what the user is interested in.

Abstract: Techniques are described for providing automated recommendations of real-world locations, such as businesses, for users to visit based at least in part on historical location-preference information. The historical location-preference information used by the recommendation system may include the historical location-preference information of the person that requests the recommendation, other people explicitly identified as participants by the requestor, and/or other people implicitly determined to be participants. The historical location-preference information may be explicit, such as “check-ins” or reviews, or implicit. Implicit participants may be identified in a variety of ways, including social network relationships and the context in which the recommendation request is submitted.

Abstract: Methods and computer readable medium for allowing a user to participate in an online conversation for a head content includes receiving a response from a user for the head content. The response from the user is geo-tagged with a location of the user. A counter indicative of number of responses received for the head content at the location is updated for the head content to include the response from the user. The responses received for the head content from the plurality of users is filtered to generate a subset of the responses related to the location of the user. The subset of responses and an interactive tag cloud for the head content is presented to the user. The subset of responses is presented in a ranked order. The tag cloud identifies one or more locations that have received greatest number of responses from the plurality of users.

Abstract: The disclosure includes a system and method for identifying advertising content and providing a user of a mobile device with a channel related to the advertising content. A recognition application receives advertising content comprising a series of frames, identifies images corresponding to the frames within the advertising content, identifies a first set of feature points from a first image of the advertising content, determines whether the first set of feature points matches a second set of feature points stored in a database, and in response to determining that the first set of feature points fails to match the second set of feature points, identifies the first image as being unique, indexes the first image and the first set of feature points and associates the first image with a channel.

Abstract: The disclosure includes a system and method for recommending an advertising campaign based on which channels are popular. A recognition application receives captured images from user devices, the captured images including information about products, identify indexed images that match the captured images, identify metadata about the captured images including information associated with products in the captured images, determine popularity of each channel based on a number of interactions that users have had with channels during a time period based on the metadata, receive an advertisement and provide on the one or more channels the advertisement based on the popularity of each channel.

Abstract: The disclosure includes a system and method for identifying discussion groups related to a video that a user would find to be interesting. A recognition application receives a captured image from a user, identifies an indexed image that matches the captured image, identifies a timestamp for the indexed image, determines a segment in the video that encompasses the timestamp, determines whether the user is interested in the segment of the video based on the timestamp and a user profile and provides suggested discussion groups to the user that are ranked based on what the user is interested in.

Abstract: The disclosure includes a system and method for grouping users into one or more categories for delivering targeted advertisements. A media application receives a captured image from a user, identifies an indexed image matching the captured image, retrieves metadata describing the captured image responsive to identifying the indexed image, generates a user profile for the user based on the metadata describing the captured image and a history of at least one previously captured image associated with the user, identifies a product in the captured image based on the metadata describing the captured image, provides the user with information about the product, updates the user profile based on how the user interacts with the product, and categorizes the user into one or more groups based on the user profile.

Abstract: The disclosure includes a system and method selecting advertisements for users based on image capturing patterns across different media. A media application receives a captured image including information about a product from a user, identifying an indexed image matching the captured image including information about the product, retrieving metadata describing the captured image including information about the product responsive to identifying the indexed image, retrieving a user profile stored for the user, identifying an advertisement from a set of advertisements based on the user profile and the metadata describing the captured image including information about the product and providing the user with the advertisement based on the user profile.

Abstract: The disclosure includes a system and method for identifying advertising content and providing a user of a mobile device with a channel related to the advertising content. A recognition application receives advertising content comprising a series of frames, identifies images corresponding to the frames within the advertising content, identifies a first set of feature points from a first image of the advertising content, determines whether the first set of feature points matches a second set of feature points stored in a database, and in response to determining that the first set of feature points fails to match the second set of feature points, identifies the first image as being unique, indexes the first image and the first set of feature points and associates the first image with a channel.

Abstract: The disclosure includes a system and method for performing actions using visual search. The system includes a controller, an image recognition module and a polling engine. The controller receives data describing an image that is associated with a poll from a first user. The image recognition module identifies the image as being an option that is selected by the first user for the poll. The polling engine associates the option selected by the first user and a first set of demographic information describing the first user with the poll. The polling engine provides the first user with polling data that describes one or more voting results for the poll. The polling data includes a second set of demographic information associated with one or more second users that provided an answer to the poll.

Abstract: Techniques are described for providing automated recommendations of real-world locations, such as businesses, for users to visit based at least in part on historical location-preference information. The historical location-preference information used by the recommendation system may include the historical location-preference information of the person that requests the recommendation, other people explicitly identified as participants by the requestor, and/or other people implicitly determined to be participants. The historical location-preference information may be explicit, such as “check-ins” or reviews, or implicit. Implicit participants may be identified in a variety of ways, including social network relationships and the context in which the recommendation request is submitted.