Day: February 5, 2020

Arrival of a Train at La Ciotat is one of the most famous films in cinema history. Shot by French filmmakers Auguste and Louis Lumière, it achieved an unprecedented level of quality for its time. Some people regard its commercial exhibition in 1896 as the birth of the film industry. An urban legend—likely apocryphal—says that viewers found the footage so realistic that they screamed and ran to the back of the room as the train approached. I’ve embedded a video of the original film above.

Of course, humanity’s standards for realism have risen dramatically over the last 125 years. Today, the Lumière brothers’ masterpiece looks grainy, murky, and basically ancient. But a man named Denis Shiryaev used modern machine-learning techniques to upscale the classic film to 21st-century video standards.

The result is remarkable. Watching the upscaled version makes the world of our great-great-great-grandparents come to life. Formerly murky details of the train, the clothing, and the faces of the passengers now stand out clearly.

How did Shiryaev do it? He says he used commercial image-editing software called Gigapixel AI. Created by Topaz Labs, the package allows customers to upscale images by up to 600 percent. Using sophisticated neural networks, Gigapixel AI adds realistic details into an image to avoid making it look blurry as it’s scaled up.

As the name implies, neural networks are networks of neurons—mathematical functions that transform a set of input values into an output value. The key feature of neural networks is that they can be trained: if you have a bunch of example inputs whose “correct” outputs are known, you can tune the parameters of the network to make it more likely to produce correct answers. The hope is that this training will generalize—that once you’ve trained it to produce the right answer for inputs the network has seen before, it will also produce good answers for inputs it hasn’t seen, too.

To train a network, you need to have a database of examples where the right answer is already known. Sometimes AI researchers have to hire human beings to produce these right answers by hand. But for a image upscaling, there’s a convenient shortcut: you start with high-resolution images and downsample them. The low-resolution images become your inputs and the high-resolution originals serve as the “correct” answer the network is aiming to produce.

“A neural network analyzes thousands of photo pairs to learn how details usually get lost,” Topaz Labs explains on their product page for Gigapixel AI. “The algorithm learns to ‘fill in’ information in new images based on what it has learned, effectively adding new detail to your photo.”

Show the neural network a low-resolution image of a face and it will figure out that it’s a face and fill in the right details for the subject’s eyes, nose, and mouth. Show the neural network a low-resolution brick building and it will add a suitable brick pattern in the high-res version.

An obvious next step would be to colorize the video. Neural networks can do that, too, using the same basic technique: start with a bunch of color photos, convert them to black and white, and then train a neural network to reconstruct the color originals.

I dropped a frame from Shiryaev’s video into the Colorize Images app for Android, which uses machine learning to automatically colorize images. As you can see, it does a pretty good job, correctly inferring that trees should be green, gravel should be a brownish color, and that men’s coats should be black. I would love to see someone with more time and better tools colorize Shiryaev’s upscaled version of the Lumière Brothers’ classic.

Everybody loves bubbles, regardless of age—the bigger the better. But to blow really big, world-record-scale bubbles requires a very precise bubble mixture. Physicists have determined that a key ingredient is mixing in polymers of varying strand lengths, according to a new paper in Physical Review Fluids. That produces a soap film able to stretch sufficiently thin to make a giant bubble without breaking.

Bubbles may seem frivolous, but there is some complex underlying physics, and hence their study has long been serious science. In the 1800s, Belgian physicist Joseph Plateau outlined four basic laws of surface tension that determine the structure of soapy films. Surface tension is why bubbles are round; that shape has the least surface area for a given volume, so it requires the least energy to maintain. Over time, that shape will start to look more like a soccer ball than a perfect sphere as gravity pulls the liquid downward (“coarsening”).

Bubbles and foams remain an active area of research. For instance, in 2016, French physicists worked out a theoretical model for the exact mechanism for how soap bubbles form when jets of air hit a soapy film. They found that bubbles only formed above a certain speed, which in turn depends on the width of the jet of air. If the jet is wide, there will be a lower threshold for forming bubbles, and those bubbles will be larger than ones produced by narrower jets, which have higher speed thresholds. That’s what’s happening, physics-wise, when we blow bubbles through a little plastic wand: the jet forms at our lips and is wider than the soapy film suspended within the wand.

In 2018, we reported on how mathematicians at New York University’s Applied Math Lab had fine-tuned the method for blowing the perfect bubble even further based on similar experiments with soapy thin films. They concluded that it’s best to use a circular wand with a 1.5-inch perimeter and gently blow at a consistent 6.9cm/s. Blow at higher speeds and the bubble will burst. Use a smaller or larger wand, and the same thing will happen.

But what about blowing gigantic bubbles or long, thin soap films that can span two stories? Justin Burton, co-author of the latest paper and a physicist at Emory University specializing in fluid dynamics, first got intrigued by the topic at a conference in Barcelona. He saw street performers producing giant bubbles about the diameter of a hula hoop and as long as a car.

He was especially intrigued by the shifting rainbow of colors on the bubbles’ surface. This effect is due to interference patterns, created when light reflects off the two surfaces of the film. For Burton, this was also an indication that the thickness of the soap was just a few microns, roughly equivalent to the wavelength of light. He was surprised that a soap film could remain intact when stretched so thin into a giant bubble and started doing his own experiments, both in the lab and his own backyard.

While perusing the open access Soap Bubble Wiki, he noticed that most of the favored recipes for bubble solution included a polymer—usually natural guar (a common thickening food additive) or a medical lubricant (polyethylene glycol).

Using those recipes as a guide, “We basically started making bubbles and popping them, and recorded the speed and dynamics of that process,” said Burton. “Focusing on a fluid at its most violent moments can tell you a lot about its underlying physics.”

The ultimate goal was to determine the perfect proportions for a bubble mixture to produce gigantic bubbles: something with a bit of stretch, but not too much, where the fluid flows a little, but not too much—in other words, the Goldilocks of bubble mixtures.

As Lissie Connors writes at Physics Buzz:

For their experiment, the researchers created various mixes of water, soap, and long-chain polymers to make their bubbles. Unfortunately, blowing a 100 m3 bubble is a poor use of lab space, and quite difficult to measure accurately, so the soap films were created using a cotton string, and the thickness was measured using infrared light. In addition to measuring the thickness, they also tracked the lifetime of each film.

Burton and his team concluded that it was the polymeric strands that were the key to producing giant bubbles, confirming the collective online wisdom. “The polymer strands become entangled, something like a hairball, forming longer strands that don’t want to break apart,” said Burton. “In the right combination, a polymer allows a soap film to reach a ‘sweet spot’ that’s viscous but also stretchy—just not so stretchy that it rips apart.”

The team also found that varying the length of the polymer strands resulted in a sturdier soap film. “Polymers of different sizes become even more entangled than single-sized polymers, strengthening the elasticity of the film,” said Burton. “That’s a fundamental physics discovery.”

You can find Burton’s giant bubble recipe in the sidebar. But be forewarned: there are some factors that can’t be controlled in a real-world setting (as opposed to Burton’s laboratory environment), like humidity levels.

DOI: Physical Review Fluids, 2020. 10.1103/PhysRevFluids.5.013304 (About DOIs).