We’ve been talking cognitive technology solutions lately. Remember that cognitive technologies are the tools that pump artificial intelligence (AI) into business. To help us get to know what we can really do with futuristic AI, we’re looking at cognitive technology examples in use in the real world. Last week we looked at cognitive technology in brick-and-mortar retail. Now let’s turn to that industry that is always fantasizing about machine takeovers: Hollywood. No, it’s not onscreen inventions we’re talking about. Let’s look at how cognitive technology is changing the way that Hollywood does business.
Movie Ads: Peppered Not Sprayed
Legendary Entertainment’s Matthew Marolda recently talked about the television and movie studio’s updated analytics plans. Marolda described the old advertising regime as one known as “spray and pray”. Studios used to try and guess at what would be most enticing to big groups and just unleash a huge number of ads on the public, hoping that seeing ads everywhere would bring people to a movie.
Now, the studio can use a whole host of cognitive technology solutions to deep-dive into understanding individual viewers. Marolda describes using “use machine learning, neural networks, [and] computer vision” to figure out how likely it is that a single viewer is interested in their content. That way the studio can target advertising to the right viewers with the message that is most engaging to them. Check out the details over at ZDNet.
Position Opening: Seeking AI Movie Trailer Production Assistant
Studio 20th Century Fox learned what a good production assistant IBM’s Watson could be when putting together a trailer for an upcoming movie, Morgan. The machine learning program was first trained to recognize which scenes of a movie are most central to the movie’s genre, in this case the horror genre. To do this, the machine learning program studied visuals and audio from 100 different horror movies. It could then look at the movie Morganand choose the top scenes related to the genre—the scary, suspenseful scenes.
So Watson saved anyone from having to manually sort through each scene in the movie to find the best options. After the scenes were chosen, the trailer’s editor could just focus on deciding which ones should be used and how the trailer should come together. How’d the Watson team do? Follow the link to the trailer and you be the judge.
Pay Attention to What’s Behind the Big Screen
Movies studios are using a lot of metadata in their filming these days. Tagging film with information like which people and products are in shots or filming location help the studio. They can keep track of budgets and progress, and can use the information along with viewer reactions to get a better sense for marketing decisions.
That can be a lot of manual tagging to do. What if, as IBM’s Josie Baik wonders, machine learning could do a lot of the heavy lifting instead? Cognitive technology solutions like image and facial recognition could be used to tag scene locations, actors, and product placements. Audio recognition could be used to categorize scenes by dialogue. Another AI assistant position seems likely to be in the making.