Last Sunday, October 29, 2019, the Internet turned 50 years old.
We’ve grown from the 1970 topology:
to this in 2019:
Okay, here’s a real representation of the Internet.
What’s next? The Interplanetary Internet of course.
My take: It’s important to note that the World Wide Web is not the same thing as the Internet. (The Web wouldn’t be invented for another 20 years!) The Internet is the all-important backbone for the numerous networking protocols that traverse it, http(s) being only one.
Insta360 has released the world’s smallest action camera, the GO. It is so small it’s potentially a choking hazard.
They call it the “20 gram steady cam.”
Here are some specs:
Some sample footage:
See some product reviews.
You can buy it now for $270 in Canada.
My take: this is too cool! My favourite features are the slow motion and the barrel roll you can add in post. This technology sparks lots of storytelling ideas!
BBC Click has revealed glimpses of the virtual production techniques Jon Favreau harnessed before the “live action” Lion King was digitally animated.
The discussion of virtual production technology starts at 0:40. Details begin flowing about the Technicolor Virtual Production pipeline at 1:38.
Director Favreau explains further at 8:01 below:
My favourite line is: “We’d move the sun if we had to.”
Here’s Technicolor’s pitch for virtual production:
More here.
My take: Am I the only one that thinks it’s absurd for photo-realistic animals to talk and sing? I can buy the anthropomorphism in most animation, as the techniques they use are suitably abstracted, but this just looks too real. Maybe thought balloons?
In the week that FaceApp went viral, Mauro Martino has updated AIportraits to convert your photos into fine art.
This web-based app uses an AI GAN trained on 54,000 fine art paintings to “paint” your portrait in a style it chooses.
Mauro explains:
“This is not a style transfer. With AI Portraits Ars anyone is able to use GAN models to generate a new painting, where facial lines are completely redesigned. The model decides for itself which style to use for the portrait. Details of the face and background contribute to direct the model towards a style. In style transfer, there is usually a strong alteration of colors, but the features of the photo remain unchanged. AI Portraits Ars creates new forms, beyond altering the style of an existing photo.”
Some samples:
He continues:
“You will not see smiles among the portraits. Portrait masters rarely paint smiling people because smiles and laughter were commonly associated with a more comic aspect of genre painting, and because the display of such an overt expression as smiling can seem to distort the face of the sitter. This inability of artificial intelligence to reproduce our smiles is teaching us something about the history of art. This and other biases that emerge in reproducing our photos with AI Portraits Ars are therefore an indirect exploration of the history of art and portraiture.”
My take: This is a lot of fun! I would love to be able to choose the “artist” though, rather than let the AI choose, based on the background. One thing that does NOT work is to feed it fine art; I tried the Mona Lisa and was terribly disappointed!
This week the BBC celebrated the 1000th episode of their technology magazine show Click with an interactive issue.
Access the show and get prepared to click!
One of the pieces that caught my eye was an item in the Tech News section about interactive art, called Mechanical Masterpieces by artist Neil Mendoza.
The exhibit is a mashup of digitized high art and Rube Goldberg-esque analogue controls that let the participants prod and poke the paintings. Very playful! I’ve scoured the web to find some video. This is Neil’s version of American Gothic:
Getting ready for the weekend with another piece from Neil Mendoza’s Mechanical Masterpieces, part of #ToughArt2018. pittsburghkids.org/exhibits/tough-art
Posted by Children's Museum of Pittsburgh on Friday, September 28, 2018
And here is his version of The Laughing Cavalier:
#ToughArt18 is now open! In @neilmendoza’s Mechanical Masterpieces, you can interact with 5 classic art images in ways you never imagined! Stay tuned to see previews of the 2 other #ToughArt pieces. pic.twitter.com/pNOj2hUNIk
— Childrens Museum Pgh (@PghKids) September 13, 2018
Check out Neil’s latest installation/music video.
My take: I love Click and I love interactive storytelling. But I’m not sure the BBC’s experiment was entirely successful. What I thought was missing was an Index, a way to quickly jump around their show. For instance, it was tortuous trying to find this item in the Tech News section. Of course, Click is in love with their material and expects viewers to patiently lap up every frame, even as they click to choose different paths through the material. But it’s documentary/news content, not narrative fiction, and I found myself wanting to jump ahead or abandon threads. On the other hand, my expectations of a narrative audience looking for A-B interactive entertainment is that they truly are motivated to explore various linear paths through the story. And an Index would reveal too much of what’s up ahead. But I wonder if that’s just me, as a creator, speaking. Perhaps interactive content is relegated to the hypertext/website side of things, versus stories that swallow you up as they twist and turn on their way to revealing their narratives.
Scientists working at Stanford University, the Max Planck Institute for Informatics, Princeton University and Adobe Research have developed a technique that synthesizes new video frames from an edited interview transcript.
In other words, soon we’ll be able to alter speech in video clips simply by typing in new words:
“Our method automatically annotates an input talking-head video with phonemes, visemes, 3D face pose and geometry, reflectance, expression and scene illumination per frame. To edit a video, the user has to only edit the transcript, and an optimization strategy then chooses segments of the input corpus as base material. The annotated parameters corresponding to the selected segments are seamlessly stitched together and used to produce an intermediate video representation in which the lower half of the face is rendered with a parametric face model. Finally, a recurrent video generation network transforms this representation to a photorealistic video that matches the edited transcript.”
Why do this?
“Our main application is text-based editing of talking-head video. We support moving and deleting phrases, and the more challenging task of adding new unspoken words. Our approach produces photo-realistic results with good audio to video alignment and a photo-realistic mouth interior including highly detailed teeth.”
Read the full research paper.
My take: Yes, this could be handy in the editing suite. But the potential for abuse is very concerning. The ease of creating Deep Fakes by simply typing new words means that we would never be able to trust any video again. No longer will a picture be worth a thousand words; rather, one word will be worth a thousand pixels.
Ivan Mehta reports in The Next Web that Samsung’s new AI can create talking avatars from a single photo.
Egor Zakharov, Aliaksandra Shysheya, Egor Burkov and Victor Lempitsky of the Skolkovo Institute of Science and Technology and the Samsung AI Center, both in Moscow, Russia, envisioned a system that…
“…performs lengthy meta-learning on a large dataset of videos, and after that is able to frame few- and one-shot learning of neural talking head models of previously unseen people as adversarial training problems with high capacity generators and discriminators. Crucially, the system is able to initialize the parameters of both the generator and the discriminator in a person-specific way, so that training can be based on just a few images and done quickly, despite the need to tune tens of millions of parameters.”
But why did the researchers set out to do this?
They wanted to make better avatars for Augmented and Virtual Reality:
“We believe that telepresence technologies in AR, VR and other media are to transform the world in the not-so-distant future. Shifting a part of human life-like communication to the virtual and augmented worlds will have several positive effects. It will lead to a reduction in long-distance travel and short-distance commute. It will democratize education, and improve the quality of life for people with disabilities. It will distribute jobs more fairly and uniformly around the World. It will better connect relatives and friends separated by distance. To achieve all these effects, we need to make human communication in AR and VR as realistic and compelling as possible, and the creation of photorealistic avatars is one (small) step towards this future. In other words, in future telepresence systems, people will need to be represented by the realistic semblances of themselves, and creating such avatars should be easy for the users. This application and scientific curiosity is what drives the research in our group.”
Read their research paper.
My take: surely this only means more Deepfakes? The one aspect of this that I think is fascinating is the potential to bring old paintings and photographs to life. I think this would be a highly creative application of the technology. With which famous portrait would you like to interact?
The Black Mirror team have handed Netflix a major win with interactive movie ‘Bandersnatch.’
Netflix has dabbled with interactive titles before, but only for kids. This outing is definitively all-grown-up with drugs, madness and violence.
As summarized on SYFYwire:
“To experience the film, viewers begin watching it like any other program or movie. But as Bandersnatch moves along, a series of choices appear on the screen, roughly every few minutes or so. Using a remote control, console controller, or keyboard, viewers make the decisions for the story’s protagonist, sending the narrative off in any number of new directions.”
Over five hours of completed scenes were shot.
“The branching narratives have been developed through a new Netflix software called Branch Manager, which can also allow viewers to exit and start all over again if they choose. The project is supported on most TVs, game consoles and either Android or iOS devices as long as they’re running the latest version of Netflix.”
How complex is the story map? See this picture or this summary. Warning: contains spoilers.
My take: I love this! The theme of who’s in control is perfect for an interactive movie about a programmer programming an interactive video game. The fact that the movie has been gamified (as the various endings have different point ratings) is hard to miss. Makes my first interactive musings look simple in comparison. I would love to make a Netflix Interactive Movie!
Will you still be on Santa’s Nice List in January?
LG will announce its latest CineBeam Laser 4K projector at CES 2019 in Las Vegas within two weeks.
The HU85L uses Ultra Short Throw (UST) technology that lets it project a 90-inch diagonal image when the unit is a mere 2 inches from the wall.
It can project a 120-inch image when placed 7 inches from the wall.
Here’s a review of last year’s model:
My take: Not cheap at something like $3,000 (3K for 4K?) but would mean you could dispense with a TV in your living room.
Tero Karras, Samuli Laine and Timo Aila of Nvidia have just published breakthrough work on Generative Adversarial Networks and images:
“We propose an alternative generator architecture for generative adversarial networks, borrowing from style transfer literature. The new architecture leads to an automatically learned, unsupervised separation of high-level attributes (e.g., pose and identity when trained on human faces) and stochastic variation in the generated images (e.g., freckles, hair), and it enables intuitive, scale-specific control of the synthesis. The new generator improves the state-of-the-art in terms of traditional distribution quality metrics, leads to demonstrably better interpolation properties, and also better disentangles the latent factors of variation.”
I’ve blogged about Nvidia’s GANs and image generation before, but this improvement in quality is remarkable.
If I understand it correctly, the breakthrough is applying one picture as a “style” or filter on another picture. Applying the filters in the left column to the pictures across the top yields the AI-generated pictures in the middle.
Read the scientific paper for full details.
Of course, we’ve seen something similar before. Way back in 1985 Godley & Creme released a music video for their song Cry; the evocative black and white video used analogue wipes and fades to blend a myriad of faces together, predating digital morphing. Here’s a cover version and video remake by Gayngs, including a cameo by Kevin Godley:
My take: Definitely scary. But if that’s the current state of the art, I think it means we are _not_ living in the Simulation — yet, even though Elon Musk says otherwise.