A machine learning system infers a “brand score” curve of a video across the run time for the video. The system uses a ground truth score obtained using user surveys, audio transcription of words spoken, video transcription of words displayed, type of music being played, and computer vision signals to learn a model for inferring the brand score. A given video is segmented, and a piecewise brand score for each segment is generated using the model.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.