GUEST EDITORIAL: 11 Possible Uses of AI in Cinema
Image: https://unsplash.com/photos/t1PaIbMTJIM
11 Possible Uses of AI in Cinema
by Jean Hartley
Artificial intelligence could very well create large-scale scenes in blockbusters much faster and better than the most talented editors and artists. But how will it work? And do such programs already exist? Here are 11 possible uses for AI in cinema.
1. Face Recognition
Face recognition is already used in everyday life, and AI performs this task quite accurately. The system identifies a person by facial features - if you take these key points and combine them with facial expression data in certain situations, such as during speech, you can get a realistic digital face that will speak and even react emotionally to other characters' lines and changes in circumstances.
2. Simulation of artificial worlds
Another application of AI in filmmaking could be the creation of a fully digital universe. In one form or another, this technology has been used in video games since 1991, when Sid Meyer's Civilization came out. Nearly three decades have passed since then, and the AI-created worlds have improved markedly, peaking in 2016 with the computer generated flora and fauna of 256 different galaxies in No Man's Sky. But what does this mean for the film industry? If the script says something like "a futuristic dystopian world with buildings from the early 20th century," the audience won't get a repainted version of New York, but a world created by the AI based on the architecture of that era.
3. Imitation of the real world
Video game universes already use textures, lighting, and physical phenomena from the real world. In a few more years, or perhaps one or two decades, AI will be able to simulate human and animal behavior to create digital spaces that replicate reality. Nvidia, one of the leading manufacturers of computer graphics software, released the open source PhysX simulator in 2019. Thanks to such solutions, everyone can contribute to the development of this technology.
4. Creation of Deepfakes
Deepfakes are images and videos which use computer vision and similar technologies to combine faces and sound. While the technology is controversial, it already has positive results in the film industry: it resurrected Peter Cushing to play Gran Moff Tarkin in Rogue One. A Star Wars Story. And, at first glance, it turned out very convincing, although it's hard to perfectly overlay a digital image on the face of a live performer. When in the same film it was required to show the young Carrie Fisher in the role of Leia, the artists combined the image of the late actress with the play of Ingvild Dale. And it didn't work out so well - any viewer will be able to tell that it wasn't Carrie Fisher they saw in that scene. But what if the deepfake was created by a machine instead of a man? Would it be better? Probably. In this film, the artists worked with so-called motion capture technology, but what if it's no longer needed?
5. Imitation of behavior
Although this point doesn't seem like an obvious necessity, such algorithms would come in handy for filming mass scenes. Large crowds of people and animals react predictably to external stimuli, so if needed to simulate a zombie or alien attack, the computer could create fairly realistic footage.
6. Copying the stars of the past
One of the most exciting prospects is creating digital copies of the stars of the past. In the near future, the addition of deceased actors to films will no longer seem fantastic: computers will learn to analyze their playing style and generate an accurate digital model. Of course, this technology raises ethical questions, but it can be used not only for the "resurrection" of celebrities. Relevant, too, is the rejuvenation or aging of living actors to play the role of a younger or older character without heavy and uncomfortable makeup.
7. Turning text into image
Even though it's still in its infancy, Microsoft already has a tool which turns text into a visual image. Users can enter a description of something, and the program will generate an image. The result is quite precise, even if it is not very detailed. You can easily imagine that in the future these visualizations will be much more detailed and can be embedded in films.
8. Film Translation
All key films are translated for distribution abroad, but a lot is lost in dubbing, this is obvious. Besides, remarks in another language don't always coincide with the actors' lip movements. Technology can solve this problem: if you combine deepfakes with services like Microsoft's Speech Translation, you can get a program that will automatically translate the characters' speech and correct their articulation.
9. Imitation of the director's style
Of course, a lot in cinema depends on directors, and they, like actors, are not eternal. And each of them also has certain traits, which the free ai image generator can analyze and reproduce exactly. Such technology will make it possible to create a Stanley Kubrick-style film that he himself never made.
10. Modeling real objects
AI works with numbers; it has no eyes, ears, or emotions, so the images it creates can be extremely realistic. The movie Interstellar used a synthesis of AI and special effects to create a stunning image of a black hole. The picture was released in 2014, and scientists received a picture of a black hole only in 2019. The computer-generated image turned out to be strikingly similar to the real photograph.
11. Reenactment of historical events
Abraham Lincoln gave his famous Gettysburg Address long before the advent of recording devices, but it can now be reconstructed. We know what Lincoln looked like, what he said and how his contemporaries perceived his performance. All this data can be entered into the program received a digital version of the event to visually demonstrate to young people one of the most important moments in American history.
ABOUT THE AUTHOR
Jean Hartley is a professional content writer who has helped students for over 6 years. She successfully works for a write my paper service and also manages remote projects. Jean has her own blog on YouTube where she shares her knowledge.
