[LINK] Patent Drops

[Stephen Loosley]

Patent Drop: October 14, 2024 https://www.thedailyupside.com/newsletter/pd-october-14-2024/
 
 
Happy Monday and welcome to Patent Drop! 

Google’s patent for a streamlined video-generation diffusion model highlights the costs and benefits of creating a high-quality AI movie generator. Plus: Snap may have a plan to make AR glasses lighter, and Mastercard might use blockchain to straighten out transaction records.


ARTIFICIAL INTELLIGENCE

Google Goes Hollywood

The tech industry is in a mad dash to make generative AI bigger and better, and Google might be looking to stake its claim in IP before it’s too late. 

The company wants to patent a system for “generative videos using diffusion models.” A diffusion model refines data over multiple iterations, making these models a good fit for generating image and audio data.

Google claims its system can create longer videos of “higher perceptual quality” than other systems, which typically lack “temporal coherence between the input video and the generated frames.” 

Unlike conventional systems, which generate videos by predicting frames one at a time, this tech models entire videos using “3D video architecture,” a process that treats a video as a three-dimensional structure rather than a flat frame, and models multiple frames simultaneously within that structure. This makes videos more natural-looking and coherent.

Google notes that this model is trained on both images and video, giving it a wider pool of high-quality, diverse data to learn from, and improving its performance. Google claims this training method is more computationally efficient, as it helps a model reach its peak performance more quickly. This tech also saves energy with simpler calculations and better memory storage techniques, Google notes. 

Tons of companies are racing to build an AI video-generation engine that actually works. OpenAI has been developing a text-to-video model, Sora, which it began teasing back in February. Earlier this month, Meta announced Movie Gen, which can create clips from text prompts and videos from custom images. (Neither of these models are publicly available.)

In August, TikTok’s parent company ByteDance launched a video-generation app called Jimeng AI for users based in China. And Google, of course, isn’t left out of the fun: The company unveiled Veo, its own AI video generator, which it will bring to YouTube Shorts in 2025 to generate clips up to six seconds long. 

Why are these companies so interested in building high-powered content generators? While at least part of the answer is simply to see if they’re capable, a larger reason could be to draw in more creators and larger audiences, said Thomas Randall, advisory director at Info-Tech Research Group. 

“They're all competing to get the largest audience, because that's where they'll get enough data to use for advertisement revenue. That's a huge source of income for them,” Randall said. Because of this, making these experiences seamless and engaging could present major revenue opportunities. 

However, like any AI model, these video generators cost an astronomical amount of resources to create, said Randall, with likely “billions of dollars in total” going toward power, data center demands, and manpower for research. Dollar amount aside, access to high-quality data can also mean the difference between realistic outputs and uncanny valley ones. 

“To make content useful and valuable, it needs to have the right kind of data going into it,” said Randall. 

SOCIAL MEDIA

Snap Lightens AR

Snap may want to lighten the load on its AR glasses. 

The social media firm is seeking to patent “external computer vision of an eyewear device.” Snap’s tech relies on external devices for motion tracking and processing of AR experiences for higher-quality interactions and reduced computational needs. 

Interaction with sensors in typical smart glasses is “not very intuitive and has a very steep learning curve,” Snap said. “Also, performing image processing to detect hand gestures involves multiple machine learning models, which consumes a great deal of hardware resources of the smart glasses.” 

Snap’s tech gives the job of movement and gesture detection to a user’s smartphone. The user’s phone and the glasses are in constant communication, so the visuals in the eyewear correspond with the users’ movements and interactions with AR objects in their vicinity. 

In the patent’s primary example, a user would set up their phone so that its front camera captures their movements. Those movements would then be mimed in real time by an AR avatar pictured in the user’s display.

Snap isn’t the only company that has sought to lighten the load on AR headsets. Meta has filed a number of patents that aim to move processing and tracking responsibilities to external devices. For any company developing AR, a sleeker, longer-last form factor is the “end goal,” said Jake Maymar, AI strategist at The Glimpse Group.

“If it costs almost no energy to run the actual glasses and everything is offset to an external phone, you can have very lightweight batteries … and the cost of manufacturing the glasses can be very inexpensive,” he said.

Cost and wearability remain the two biggest hurdles in making true AR glasses a reality, said Maymar. Both Snap and Meta unveiled their latest smart glasses last month, and both face those issues. While Snap’s new glasses in the Spectacles line are more powerful than previous models, they’re also quite large and bulky. Meanwhile, Meta’s Project Orion AR glasses are slightly sleeker, but reportedly cost $10,000 per unit to make. 

Creating a lighter and cheaper headset may only be half the battle, said Maymar. To get people to actually use these devices, they need to seamlessly integrate into a user’s existing tech and everyday activities, he said. This is where a company like Apple — which only jumped into artificial reality at the start of this year with the Vision Pro — may have a leg-up. 

Even if Apple isn’t first, if the company puts out AR glasses, Maymar said, “ultimately, Apple is going to win. Just think: How many people have iPhones? That's a pretty clear indicator.”


BLOCKCHAIN

Mastercard’s Record Keeper

Mastercard wants to set the record straight.

The credit card giant filed a patent application for “transaction processing with complete cryptographic auditability.” This adds a layer of trackability and transparency to transactions using a blockchain ledger. 

Mastercard’s tech aims to give a user or business auditability of all their transactions in the event of a dispute, as well as “eliminate or strongly mitigate the possibility of the record being tampered with.” 

When a transaction is requested between two parties, a “third party” moderator is used to verify the details, and a digital signature is generated as a means of verification. The server that’s processing the transaction then creates its own digital signatures, adding a layer of cryptographic verification to ensure the transaction’s validity.

Once the transaction is confirmed, both the digital signatures and the details of the transaction itself are stored on a blockchain, ensuring that the record is secure, tamper-proof and immutable. 

Mastercard is far from new to the blockchain landscape. The company has filed tons of patents that use blockchain both for cryptocurrency and other use cases, such as ticket tracking and fraud detection. It’s following pace with fintech and finance firms like PayPal, Visa, and JPMorgan Chase, all of which have filed patents for blockchain-related tech. 

This patent plays to blockchain’s biggest strengths — immutability, transparency, and security — without being explicitly related to cryptocurrency, said Jordan Gutt, Web 3.0 lead at The Glimpse Group. 

“It’s using the best of blockchain technology,” said Gutt. “Mastercard wanted to find a use case for blockchain that could actually help them in solving real-world problems.” 

This invention alone presents several potential use cases, Gutt said. In one example, the filing discusses its capabilities in real estate transactions, such as escrow payments, which are “usually slow, expensive, and error-prone,” Gutt said. But this tech can provide an additional layer of trust and transparency in the purchase of any hot-ticket item, he said. 

And with the amount of purchase disputes that a company like Mastercard deals with on a regular basis, an immutable ledger could help merchants and customers easily track where payments went awry. “This patent could help either party without extra time and effort,” he said. 

“It seems like a great use case, and in the future, could save them money on a lot of different auditing tasks and reduce fraud,” said Gutt

--