Provided are systems and a method for photorealistic real-time face reenactment. An example method includes receiving a target video including a target face and a scenario including a series of source facial expressions, determining, based on the target face, one or more target facial expressions, and synthesizing, using the parametric face model, an output face. The output face includes the target face. The one or more target facial expressions are modified to imitate the source facial expressions. The method further includes generating, based on a deep neural network, a mouth region and an eyes region, and combining the output face, the mouth region, and the eyes region to generate a frame of an output video.

A system and method for providing dynamic pay-for-placement advertisements via graphics-enabled email that generates a display of advertisements when the email newsletter is opened so the advertisements displayed are based on rankings at the time the email is opened instead of when the email was generated and transmitted. In one embodiment, a graphical-content email having one or more embedded advertisement image references is provided to one or more email recipients. The advertisement image reference, in one embodiment, may include query string parameters indicating the context of the image reference and/or portion of the image reference (i.e., identifying the image reference as being part of a particular newsletter email), a position of the image reference in the email display, and the like. A URL reference also may be included with each advertisement image reference (.e.g., one URL for each advertisement portion of the image to be retrieved by the advertising image reference).

Systems and methods are disclosed for online distribution of content based on a user sentiment index. The method may include receiving, over a network and from a user device, one or more user generated inputs and calculating the user sentiment index based on the one or more user generated inputs. The method may also include receiving, over the network, from a content or advertising provider, instructions on publishing content or advertising to a webpage based on the calculated user sentiment index, and publishing content for display on user devices over the network based on a comparison of the calculated user sentiment index and the received instructions.

Off-policy evaluation of a new target policy is performed using historical data gathered based on a previous logging policy to estimate the performance of the target policy. An estimator may be used, wherein either a quality-based estimator or a quality-agnostic estimator is used to weight the difference between an observed reward in the historical data and an estimated reward generated by the target policy. A quality-agnostic estimator may be used to evaluate an importance weight according to a threshold. In such examples, when the importance weight exceeds the threshold, the quality-agnostic estimator clips the importance weight at the threshold, thereby providing an fixed upper bound irrespective of the quality of the reward predictor. In other examples, a quality-based estimator is used, in which an upper bound incorporates the quality of the reward predictor in order to modify an importance weight used by the estimator.

A software application that is configured to broker interactions intermediate a plurality of businesses and clients wherein the software application will attempt to facilitate the consumption of excess capacity of the businesses. The present invention is executed via a software platform controlled by an operator wherein the operator enrolls a plurality of businesses and clients. The businesses establish and publish a capacity calendar wherein the capacity calendar provides information as to the availability of the product or service of the business. The clients establish a preferred business list either by identification of a particular business or category. The software application provides cross-referencing of the product or service availability and notifies clients upon availability of a service or product. The operator of the present invention further provides reporting and advertising for businesses enrolled with the present invention. A reward system is associated with purchases made through the present invention.

A method, which comprises determining a logical identifier used for communication over a network portion managed by a service provider and transmitting to a computing apparatus adapted to effect online transactions involving a payer and a payee an indication of the service provider having agreed to act as the payer for at least one online transaction requested by a requesting device that uses the logical identifier. Also, a method, which comprises determining a logical identifier used to identify a device during a request for an online transaction; identifying, based on the logical identifier, a third party that has agreed to act as a payer for the online transaction; providing a user of the device with an opportunity to confirm the third party as the payer for the online transaction; and completing the online transaction based on input from the user.

Provided is a computer-implemented method for predicting payment transactions using a machine learning technique that includes receiving transaction data, generating a categorical transaction model based on the transaction data, determining a plurality of prediction scores including determining, for one or more users, a prediction score in each merchant category of a plurality of merchant categories for each predetermined time segment of a plurality of predetermined time segments, where a respective prediction score includes a prediction of whether a user will conduct a payment transaction in a merchant category at a time associated with a predetermined time segment associated with the respective prediction score, determining a recommended merchant category and a recommended predetermined time segment of at least one offer, generating the at least one offer, and communicating the at least one offer to the one or more users. A system and computer program product are also disclosed.

A method (100) of controlling playout of advertisement content during live video streaming at an end-user terminal comprising steps of: receiving (110) advertisement content from an advertisement server; receiving (112) live streamed video content from a content delivery network and playing out the video content; obtaining (114) at least one of image features and audio features of the video content during playout; calculating (116) a content importance rating of video content to be played out during a prediction time window based on said features; and postponing (118) playout of advertisement content scheduled to be played out during the prediction time window if the calculated content importance rating for the prediction time window is above a threshold value.

A machine-learning algorithm is trained with features relevant to a modeled set of input directed to patterns of activities specific to a given behavior. The trained algorithm is also trained on success and failures of remediation actions that change or do not change the given behavior. The trained algorithm is then provided the modeled set of input at predefined intervals of time and supplies as output expected deviations/changes that are predicted for the given behavior along with an indication as to whether the remediation actions are likely to prevent or change the expected behaviors.

A content provider specifies a content campaign along with a desired metric to achieve for the campaign, such as a budget. For each opportunity to present content to a user, an online publishing system selects content to a group of users using a selection process. To achieve the desired metric for the content campaign, the publishing system determines a probability of participating in the selection process with content from the content campaign based on feedback about the current performance of the campaign. For example, the system compares the opportunities passed and the budget spent for the content campaign to a current time and budget expenditure, and then it calculates an equilibrium probability that will achieve the desired budget at the end of the campaign. The system then determines whether to participate in a selection process for each new opportunity to present content using this adjusted equilibrium probability.