20th Century Fox is working on an AI to predict upcoming movie performances. Read more

In what could be the next big step in widening the range of implementation of Artificial Intelligence (AI), researchers at film studio 20th Century Fox are said to be working on a deep learning system that would be able to predict the performance of a movie in the market by analysing its trailer. 

The new AI would be able to link the visual elements of a trailer (such as faces, landscapes, lighting and colours) and the performance of a film for certain demographics. 

The research behind the innovation found that a trailer with more number of people talking and warm colours might appeal to a different group than the ones with dull colours and more visual dependence.  

The method already seems to be working in the real world when used on existing movies. The company is also said to be testing the new method on its upcoming movies. 

"The trained model not only makes accurate attendance and audience prediction for existing movies, but also successfully profiles new movies six to eight months prior to their release." the paper said.

However, the study and its applications come with a handful of setbacks. For instance, the AI does not capture the temporal information of the screen which involves representing and reasoning spatial-temporal knowledge in mind. The applied goal—on the computing side—involves developing high-level control systems of robots for navigating and understanding time and space. Additionally, it does not distinguish between the video and the text description in order to achieve a complete sense of the movie.

A Collaborative Filtering Model (CFM) that was developed for the research was used to predict customer behaviour six to eight months after the trailer was released.

READ: High Alert office Snoozers! Here Comes The Foe To Your Short-naps

This new project could be useful for entertainment companies which could make a target-oriented preview of the movie for its audience that would attract a larger numbers for its movie. This would be good news for all the entertainment companies in the age of streaming where movies are easily available on platforms such as Netflix or Amazon Prime. 

READ: Google Might Have Accidentally Leaked Its Biggest Pixel Upgrade Ever

"Movie-streaming differs from moviegoing. It is cheaper and reversible (one can stop the movie), and there is less risk involved. Casual streamers are not penalized for making mistakes when drawing from a broader catalog as casual moviegoers are if they do not like the movie" the paper suggested.