Automated Content Evaluation Using a Predictive Model

Number: 20200151524
Published: 2020-05-14
Filed: 2019-05-08
Assignee: Disney Enterprises, Inc.
Inventors: Lombardo; Salvator D., Segalin; Cristina, Chen; Lei, Navarathna; Rajitha D., Mandt; Stephan Marcel
CPC: G06V10/82; G06V30/2504; G06N3/09; G06V10/7715; G06N3/049; G06F18/254; G06V40/168; G06F18/241; G06N3/08; G06N3/045; G06N3/0464; G06V20/52; G06F18/2115; G06V40/165; G06N3/047; G06Q30/02; G06N3/0455; G06F18/2148; G06N3/088; G06V10/809
Verdict: Set aside content-quality predictive model, business
Source: Google Patents · FreePatentsOnline

Abstract

There are provided systems and methods for performing automated content evaluation. In one implementation, the system includes a hardware processor and a system memory storing a software code including a predictive model trained based on an audience response to training content. The hardware processor executes the software code to receive images, each image including facial landmarks of an audience member viewing the content during its duration, and for each image, transforms the facial landmarks to a lower dimensional facial representation, resulting in multiple lower dimensional facial representations of each audience member. For each of a subset of the lower dimensional facial representations of each audience member, the software code utilizes the predictive model to predict one or more responses to the content, resulting in multiple predictions for each audience member, and classifies one or more time segment(s) in the duration of the content based on an aggregate of the predictions.

Background

BACKGROUND

Media content in a wide variety of formats is consistently sought out and enjoyed by consumers. Nevertheless, the popularity of a particular item or items of media content, such as a movie, television (TV) series, or a particular TV episode, for example, can vary widely. One approach to evaluating the potential desirability of media content is to use an audience as a focus group to help understand whether a TV episode, for example, is or will be successful.

Traditional approaches used in audience analysis typically require annotated data in order to identify certain expressions in the audience members faces during screening of the content. However, those traditional approaches require extensive manual annotation of large datasets, rendering them expensive and time consuming to prepare. Due to the resources often devoted to developing new content, the accuracy and efficiency with which the desirability of such content to consumers can be evaluated has become increasingly important to producers, owners, and distributors of media content.SUMMARY

There are provided systems and methods for performing automated content evaluation using a predictive model, substantially as shown in and/or described in connection with at least one of the figures, and as set forth more completely in the claims.

Claims

1. A content evaluation system comprising: a hardware processor and a system memory; a software code stored in the system memory, the software code including a predictive model trained based on an audience response to a training content; the hardware processor configured to execute the software code to: receive a plurality of images each including a plurality of facial landmarks of one of a plurality of audience members viewing a content during a duration of the content; for each of the plurality of images, transform the plurality of facial landmarks to a lower dimensional facial representation, resulting in a plurality of lower dimensional facial representations of the one of the plurality of audience members; for each of a subset of the plurality of lower dimensional facial representations of the one of the plurality of audience members, utilize the predictive model to predict one or more responses to the content, resulting in a plurality of predictions for each of the plurality of audience members; and classify at least one time segment in the duration of the content based on an aggregate of the plurality of predictions for the plurality of audience members. 11. A method for use by a content evaluation system including a hardware processor and a system memory storing a software code including a predictive model trained based on an audience response to a training content, the method comprising: receiving, by the software code executed by the hardware processor, a plurality of images each including a plurality of facial landmarks of one of a plurality of audience members viewing a content during a duration of the content; for each of the plurality of images, transforming, by the software code executed by the hardware processor, the plurality of facial landmarks to a lower dimensional facial representation, resulting in a plurality of lower dimensional facial representations of the one of the plurality of audience members; for each of a subset of the plurality of lower dimensional facial representations of the one of the plurality of audience members, predicting, by the software code executed by the hardware processor and using the predictive model, one or more responses to the content, resulting in a plurality of predictions for each of the plurality of audience members; and classifying, by the software code executed by the hardware processor, at least one time segment in the duration of the content based on an aggregate of the plurality of predictions for the plurality of audience members.