AI is already being used in post production, says Julian Nelson, the head of Post Production and co-founder of Residence Pictures.
The media and entertainment (M&E) industry is no stranger to AI. In fact, the use of machine learning (ML) and deep learning algorithms is becoming increasingly prevalent within M&E.
As you may know, both ML and DL are useful for recognizing patterns within large datasets. In turn, this can be used to inform decision-making processes that require a level of precision that humans alone cannot achieve on their own.
For example, ML helps create concept art in movie making or television production by automatically suggesting possible aesthetic choices based on existing data sets containing thousands or millions of images from other movies or TV shows. This technology has been used so far most prominently in Hollywood’s blockbuster film Avengers: Endgame—which grossed $2 billion worldwide—to generate more than 100 million facial expressions for its characters based on their previous appearances across different Marvel movies over time!
Facial recognition is a sophisticated AI technology that can be used for many purposes. It can also be used for nefarious reasons. Facial recognition tech enables us to identify people based on their facial features, which makes it helpful for law enforcement and marketing purposes, but also presents a privacy threat in the wrong hands.
In recent years, this technology has been used by video games and social media platforms like Facebook to identify users’ photos in order to suggest new friends or suggest content they might enjoy seeing next time they log on. However, governments have started using facial recognition tech as well: police departments across the country have adopted such systems both as tools for solving crimes and more insidiously as tools of surveillance (especially when combined with other technologies).
Deep learning is a type of machine learning that uses neural networks. Deep learning uses multiple layers of artificial neurons to process data and make predictions. Neural networks were first developed in 1958 by Warren McCulloch and Walter Pitts, who used them to create an algorithm capable of recognizing specific numbers. More recently, deep learning has been applied to image recognition, speech recognition, natural language processing (NLP), object detection (for example in self-driving cars), and other applications.
Deep learning has also enabled new advancements in facial recognition technology. For example: Facebook’s DeepFace system can recognize faces with 97% accuracy; Google’s Inception-v3 model can identify objects with 98% accuracy; Yann LeCun’s Resnet50 model can recognize handwritten digits with 99% accuracy; Microsoft Research Asia’s ResNet V2 model can recognize street signs from satellite images 90% of the time; OpenAI Five defeated five human experts at Dota 2 after training for only two weeks.
I’m going to pause there for a sec, full disclosure this whole article so far has been generated using AI along with the accompanying image. I used copy.ai to generate the text by entering into the box: “AI Tools in Post Production”, adding key words “AI, technology, post production, editing, HDR, video, concept art, colour grading, future”. I then finished this off by asking it to deliver the article in a witty tone. I ran it through my.plag.ai which told me there was a 0% similarity with no risk of plagiarism.
The image was generated using midjourney.com. In the prompt I entered “a flowchart diagram exploring AI technology, highly detailed, hand drawn –ar 16:9”. After asking the computer to generate a couple of versions I max upscaled the image and that is what is being used here.
The text and image are unedited and not altered at all. With hardly any work from me and not much input I have created a rather convincing article on AI using AI.
Is this the future of post?
With the increasing use of AI and ML in post production, we must remain aware of its importance to us as professionals but also take advantage of what it can offer us as artists. While there are many benefits to using this technology, there will also be risks associated with its adoption. We must be vigilant in how we use it so that we can ensure our work remains authentic and true to our craft. (This last section was also generated using AI)
Share this story