Neural Networks Model Audience Reactions to Movies

Summary: A new artificial neural network can assess a viewer's reaction to movies based on patterns of facial expressions. With enough information, researchers say the ANN will be able to assess how an audience is reacting to a movie and predict an individual's response based on a few minutes of observation. Software automatically discovers patterns in facial expressions. Engineers have created a new deep-learning software capable of assessing complex audience reactions to movies using the viewer's facial expressions. Developed by Disney Research in collaboration with Yisong Yue of Caltech and colleagues at Simon Fraser University, the software relies on a new algorithm known as factorized variational autoencoders (FVAEs).