philip lelyveld The world of entertainment technology

NVIDIA ACE Enables Easier Interactive Avatars

... To deploy interactive avatars such as Violet, developers need to produce the 3D character can use computer vision, natural language understanding, AI and advanced avatar visualization tools. This has been extremely difficult up until now. While Epic Games provided a powerful tool for MetaHumans to create the avatar itself, most applications of the MetaHumans typically had them being driven by a person, akin to a puppet. The ACE platform provides artists and TD with a way to build digital humans driven by a core AI engine. This allows the avatars to effectively be independent and ‘come alive’. ...

How Omniverse ACE Brings Violet to Life 

The demo shows Violet as a fully rigged avatar with basic animation. To create Violet, NVIDIA’s team used the company’s Unified Compute Framework (UCF), a fully accelerated framework that enables developers to combine accelerated microservices into real-time AI applications. UCF helped the team build a graph of microservices for Violet that were deployed in the cloud. ...

Omniverse ACE powers the backend of interactive avatars, essentially acting as Violet’s brain. Additionally, two reference applications are built on ACE: NVIDIA Tokkio and NVIDIA Maxine. NVIDIA Tokkio leverages ACE, to make it easier to build and customize lifelike virtual assistants and digital humans. ...

Omniverse ACE also delivers microservices that enable developers to access the NVIDIA AI technology, with no coding required. Some of the latest microservices include:

• Animation AI: Omniverse Audio2Face simplifies the animation of a 3D character to match any voice-over track, helping users animate characters for games, films or real-time digital assistants.
• Conversational AI: Includes the NVIDIA Riva SDK for speech AI and NVIDIA NeMo Megatron framework for natural language processing, allowing developers to quickly build and deploy cutting-edge applications that deliver high-accuracy, expressive voices and respond in real-time. ...

See the full story here: https://www.fxguide.com/quicktakes/nvidia-ace-enables-easier-interactive-avatars/?fbclid=IwAR1Z3iZ0ATcq_GNKRCgjv8XE-25qI-O1TN3_AsI6uAI1n7zLm-vPEemdps0

HBO MAX CUTS CIGARETTE FROM ICONIC MOVIE POSTER

Last week, HBO Max, the Warner Bros.-owned TV streaming platform, cut more than just their costs — they’re cutting back on cigarettes too.

Keen observers noticed that HBO Max removed the cigarillo from the iconic movie poster from “McCabe & Mrs. Miller.” Now, McCabe is awkwardly holding up two fingers with no smoking device in hand. They also scrubbed cigarettes from several other film posters, including “The Life and Times of Judge Roy Bean,” “There Was a Crooked Man,” “Fallen Angels,” and “The Man Who Knew Too Much.” ...

... The question is not whether smoking cigarettes is bad for you. We know that it is. But smoking was nonetheless a significant aspect of daily life in American culture when these films were produced. Trying to get rid of it feels like abusing our technological capacities. Why can’t we interact with these stories for what they are and make our own judgments about them? And if this is about promoting physical health, why not place disclaimers at the front of the films that warn of the negative effects of smoking? There are other ways to discourage cigarette use than trying to pretend they weren’t a part of the movies. ...

See the full story here: https://mindmatters.ai/2022/09/hbo-max-cuts-cigarette-from-iconic-movie-poster/

IDC Tracker Sees a Long Road Ahead for Augmented Reality Headsets

Worldwide shipments of Augmented Reality (AR) headsets are forecast to decline 8.7% year over year to nearly 260,000 units by the end of 2022, according to new data from the International Data Corporation (IDC) Worldwide Quarterly Augmented and Virtual Reality Headset Tracker. While a decline for such a nascent technology may seem abrupt, volumes have been slowing in recent quarters due to the shaky financials of some of the top companies, lack of new devices, and end-user demand being fulfilled by AR on smartphones in the short term.

In the first half of 2022, the top 5 AR headset makers were Microsoft, Rokid, Shadow Creator, Mad Gaze, and RealWear. However, the landscape is expected to shift dramatically as more companies enter the market and tech giants like Apple and Meta take more of an interest in AR. Looking ahead, despite the decline in 2022, IDC anticipates plenty of growth as the market is forecast to have a 5-year compound annual growth rate (CAGR) of 70.3%, topping out at 4.1 million units by the end of 2026. ...

See the full story here: https://www.businesswire.com/news/home/20220920005157/en/IDC-Tracker-Sees-a-Long-Road-Ahead-for-Augmented-Reality-Headsets

Meta Sees VR Fitness Revolution, Will Make Its Next Headsets “Sweat-Proof”

... “The next iteration that we’re going to have of our current Quest 2 product is to make the headset more sweat proof, to have the controllers have better grips,” Shaw said. “People are using our platform to workout, and it’s not something we necessarily forecasted. We now have creators who are developing these games for people to be active. So I think gaming is going to change in many ways.” ...

Meta also partnered with the NFL around Super Bowl LVI to let fans dress their digital avatars in Rams and Bengals-branded clothing. The digital apparel was free to buy across Meta’s platforms, but the company expects to eventually have users pay for their virtual avatar’s swag. ...

See the full story here: https://sporttechie.com/louisville-slugger-hitting-science-center-uplift-labs/

Nvidia enables broader usage of AI with LLM cloud services

... Kharya explained that with a transformer, the AI model can understand which parts of a sentence, an image or even very disparate data points are relevant to each other. Unlike convolutional neural networks (CNNs), which typically look at only the immediate neighboring relationships, transformers are designed to train on more distant relationships as well, which Kharya said is very important for use cases like natural language processing (NLP). ...

The GPT-3 LLM has helped to increase awareness and adoption of LLMs for a variety of use cases, including summation and text generation. An LLM is also at the foundation of the DALL-E text-to-image generation technology. ...

The giant AI models that the service will support include a 5 billion- and a 20 billion-parameter GPT-based model, as well as one based on the Megatron 530-billion parameter LLM. As part of the service, Nvidia is also supporting prompt learning–based tuning to rapidly enable context-specific use cases. Kharya said that the NeMo LLM Service will also include the option to use both ready-made models and custom models through a cloud-based API experience.

Going a step further, Nvidia is also launching a specific LLM capability for life sciences with the BioNeMo Service. ...

See the full story here: https://venturebeat.com/ai/nvidia-enables-broader-usage-of-ai-with-llm-cloud-services/

AI Art Is Here and the World Is Already Different

... When a friend who had been experimenting with DALL-E 2 described the tool as a place to dispose of intrusive thoughts, I nodded, scrolling back in my Midjourney window to a pretty convincing take on “Joe Biden tanning on the beach drawn by R. Crumb.”

I still use Midjourney this way, but the novelty has worn off, in no small part because the renderings have just gotten better — less “strange and beautiful” than “competent and plausible.” The bit has also gotten stale, and I’ve mapped the narrow boundaries of my artistic imagination. A lot of the AI art that has gone viral was generated from prompts that produced just the right kind of result: close enough to be startling but still somehow off, through a misinterpreted word, a strange artifact that turned the image macabre, or a fully haywire conceptual interpolation. Surprising errors are AI imagery’s best approximation of genuine creativity, or at least its most joyful. TikTok’s primitive take on an image generator, which it released last month, embraces this. ...

When AI art fails a little, as it has consistently in this early phase, it’s funny. When it simply succeeds, as it will more and more convincingly in the months and years ahead, it’s just, well, automation. ...

See the full story here: https://nymag.com/intelligencer/2022/09/ai-art-is-here-and-the-world-is-already-different.html

Disney Files Patent for Augmented Reality Ride Without Glasses

This month, they have published a patent application for a visual effect system that allows theme park guests to see 3D augmented reality imagery without glasses.

The system "allows a viewer with eyes positioned at the left and right eye box positions to perceive a three-dimensional (3D) virtual object in the physical space concurrently with light from physical objects or scenery in the physical space," according to the patent filing.

In other words, if I can describe this correctly, instead of projecting left and right images that you would see through a pair of glasses as a 3D image, the two images would be projected on a transparent surface in front of your ride vehicle - that would do the job of the glasses in bringing the two images together as one, 3D image.

See the full story here: https://www.themeparkinsider.com/flume/202209/9139/

Scene-aware Sound Rendering in Virtual and Real Worlds

Abstract:Modern computer graphics applications including virtual reality (VR) and augmented reality (AR) have adopted techniques for both visual rendering and audio rendering. While visual rendering can already synthesize virtual objects into the real world seamlessly, it remains difficult to correctly blend virtual sound with real-world sound using state-of-the-art audio rendering. When the virtual sound is generated unaware of the scene, the corresponding application becomes less immersive, especially for AR. In this position paper, we present our current work on generating scene-aware sound using ray-tracing based simulation combined with deep learning and optimization.Published in: 2020 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW)

See the full story here: https://ieeexplore.ieee.org/abstract/document/9090553

Gucci built a persistent town inside of Roblox

Gucci is making its presence inside of Roblox a little more permanent. The fashion brand previously collaborated with the virtual world for more short-lived experiences, like the surreal Gucci Garden from last year, and is now releasing a more persistent space called Gucci Town. The space features a central garden that links together various areas, including a space for mini-games, a cafe, and a virtual store where players can, of course, purchase virtual Gucci gear for their Robloxavatar. (The company says the virtual outfits make use of Roblox’s new “layered clothing” tech.)

Gucci says that more than 20 million players visited the garden last year, despite it only being available for two weeks, and it expects the new space to steadily evolve over time. ...

See the full story here: https://www.theverge.com/2022/5/27/23143404/gucci-town-roblox

Antarctic virtual reality coming to Katikati

... Through VR, you’ll get to step inside Sir Edmund Hillary's hut, discover what it took to be an Antarctic explorer, and see what life was like living in the coldest place on Earth back in 1957.

This free event will be on 10am-4pm for three days only and is best suited for those aged nine years and older. Places are limited so bookings are essential - head to eventbrite.co.nz to book your session. ...

See the full story here: https://www.sunlive.co.nz/news/302922-antarctic-virtual-reality-coming-to-katikati.html