Fasten your seatbelts. NVIDIA Research is revving up a new deep learning engine that creates 3D object models from standard 2D images - and can bring iconic cars like the Knight Rider's AI-powered KITT to life - in NVIDIA Omniverse.
Developed by the NVIDIA AI Research Lab in Toronto, the GANverse3D application inflates flat images into realistic 3D models that can be visualized and controlled in virtual environments. This capability could help architects, creators, game developers and designers easily add new objects to their mockups without needing expertise in 3D modeling, or a large budget to spend on renderings.
A single photo of a car, for example, could be turned into a 3D model that can drive around a virtual scene, complete with realistic headlights, tail lights and blinkers.
To generate a dataset for training, the researchers harnessed a generative adversarial network, or GAN, to synthesize images depicting the same object from multiple viewpoints - like a photographer who walks around a parked vehicle, taking shots from different angles. These multi-view images were plugged into a rendering framework for inverse graphics, the process of inferring 3D mesh models from 2D images.
Once trained on multi-view images, GANverse3D needs only a single 2D image to predict a 3D mesh model. This model can be used with a 3D neural renderer that gives developers control to customize objects and swap out backgrounds.
When imported as an extension in the NVIDIA Omniverse platform and run on NVIDIA RTX GPUs, GANverse3D can be used to recreate any 2D image into 3D - like the beloved crime-fighting car KITT, from the popular 1980s Knight Rider TV show.
Previous models for inverse graphics have relied on 3D shapes as training data.
Instead, with no aid from 3D assets, We turned a GAN model into a very efficient data generator so we can create 3D objects from any 2D image on the web, said Wenzheng Chen, research scientist at NVIDIA and lead author on the project.
Because we trained on real images instead of the typical pipeline, which relies on synthetic data, the AI model generalizes better to real-world applications, said NVIDIA researcher Jun Gao, an author on the project.
The research behind GANverse3D will be presented at two upcoming conferences: the International Conference on Learning Representations in May, and the Conference on Computer Vision and Pattern Recognition, in June.
From Flat Tire to Racing KITT Creators in gaming, architecture and design rely on virtual environments like the NVIDIA Omniverse simulation and collaboration platform to test out new ideas and visualize prototypes before creating their final products. With Omniverse Connectors, developers can use their preferred 3D applications in Omniverse to simulate complex virtual worlds with real-time ray tracing.
But not every creator has the time and resources to create 3D models of every object they sketch. The cost of capturing the number of multi-view images necessary to render a showroom's worth of cars, or a street's worth of buildings, can be prohibitive.
That's where a trained GANverse3D application can be used to convert standard images of a car, a building or even a horse into a 3D figure that can be customized and animated in Omniverse.
To recreate KITT, the researchers simply fed the trained model an image of the car, letting GANverse3D predict a corresponding 3D textured mesh, as well as different parts of the vehicle such as wheels and headlights. They then used NVIDIA Omniverse Kit and NVIDIA PhysX tools to convert the predicted texture into high-quality materials that give KITT a more realistic look and feel, and placed it in a dynamic driving sequence.
Omniverse allows researchers to bring exciting, cutting-edge research directly to creators and end users, said Jean-Francois Lafleche, deep learning engineer at NVIDIA. Offering GANverse3D as an extension in Omniverse will help artists create richer virtual worlds for game development, city planning or even training new machine learning models.
GANs Power a Dimensional Shift Because real-world datasets that capture the same object from different angles are rare, most AI tools that convert images from 2D to 3D are trained using synthetic 3D datasets like ShapeNet.
To obtain multi-view images from real-world data - like images of cars available publicly on the web - the NVIDIA researchers instead turned to a GAN model, manipulating its neural network layers to turn it into a data generator.
The team found that opening the first four layers of the neural network and freezing the remaining 12 caused the GAN to render images of the same object from different viewpoints.
Keeping the first four layers frozen and the other 12 layers variable caused the neural network to generate different images from the same viewpoint. By manually assigning standard viewpoints, with vehicles pictured at a specific elevation and camera distance, the researchers could rapidly generate a multi-view dataset from individual 2D images.
The final model, trained on 55,000 car images generated by the GAN, outperformed an inverse graphics network trained on the popular Pascal3D dataset.
Read the full ICLR paper, authored by Wenzheng Chen, fellow NVIDIA researchers Jun Gao and Huan Ling, Sanja Fidler, director of NVIDIA's Toronto research lab, University of Waterloo student Yuxuan Zhang, Stanford student Yinan Zhang and MIT professor Antonio Torralba. Additional collaborators on the CVPR paper include Jean-Francois Lafleche, NVIDIA researcher Kangxue Yin and Adela Barriuso.
The NVIDIA Research team consists of more than 200 scientists around the globe, focusing on areas such as AI, computer vision, self-driving cars, robotics and graphics. Learn more about the company's latest research and industry breakthroughs in NVIDIA CEO Jensen Huang's k
Most recent headlines
06/10/2025
France T l visions, France's leading broadcaster, has received the 2025 EBU ...
04/09/2025
Monumental Sports & Entertainment (MSE), in collaboration with Dalet, has been a...
07/08/2025
July 8 2025, 22:30 (PDT) Tata Motors & Dolby Bring Dolby Atmos to Harrier.ev, R...
15/07/2025
The National Film and Video Foundation (NFVF), an agency of the Department of Sp...
15/07/2025
L3Harris put its CORVUS-RAVEN counter-small UAS capability into the hands of soldiers at VANAHEIM, showcasing its ability to provide passive signal detect, enha...
15/07/2025
Netflix Viewing Up 13.5% vs. May, Represents 42% of Monthly Gain for Streaming
...
15/07/2025
SAN FRANCISCO and BENGALARU, India Sangeeta Chakraborty has been named chief revenue officer at Amagi, a cloud-based software-as-a-service (SaaS) technology pro...
15/07/2025
As part of its mandate, the Advanced Television Systems Committee the U.S. organization tasked with developing advanced broadcast TV standards promotes ATSC 3.0...
15/07/2025
QuickLink, the leading global provider of multi-camera video productions and remote contributions, announces the launch of two innovative control panels Studi...
15/07/2025
Chris Kontopanos has joined leading high-quality microphone solutions manufacturer, DPA Microphones, as the company's new Regional Sales Manager for the Mid...
15/07/2025
Magnifi by VideoVerse, a global leader in AI-powered video automation, has launched a major platform upgrade built to simplify live and archival video editing f...
15/07/2025
Telestream, a global leader in media workflow technologies, will preview its latest innovations at IBC2025, Stand 7.B21. This year's showcase highlights how...
15/07/2025
IBC2025 is set to transform the onsite experience for its global attendees with the launch of a pioneering new feature: full integration of Amsterdam's GVB ...
15/07/2025
Music, sport and technology collided in spectacular fashion as ASB GlassFloor powered the debut of Beats N Buckets, Germany's first Basketball meets Hip-Ho...
15/07/2025
The Collectv, the Emmy award-winning broadcast solutions and workflows consultancy - and winner of Broadcast Tech's Team of the Year is delighted to announc...
15/07/2025
SipRadius, the expert in secure, low latency media transport, will showcase how broadcasters and media companies can take control of fragmented IP workflows at ...
15/07/2025
Appear, the global leader in live production technology, today announced its strategic partnership with beIN ASIA PACIFIC, a leading multi-platform sports media...
15/07/2025
Company marks 15 years of innovation at the show
Broadpeak, a leader in streaming and monetization at scale, will return to IBC (Hall 1, Stand F83. RAI, Amster...
15/07/2025
New software solutions extend trusted timing measurement across broadcast, venue, and professional AV applications
Hitomi Broadcast, the market leader in audi...
15/07/2025
SHENZHEN, China Telycam has added control options for its portfolio of pan-tilt-zoom (PTZ) cameras with a new plug-in for Elgato's Stream Deck family of con...
15/07/2025
MELVILLE, N.Y. Chyron said its Chyron LIVE cloud-native live production solution has successfully completed the Amazon Web Services (AWS) Foundational Technical...
15/07/2025
PHILADELPHIA The Xumo streaming platform joint venture between Comcast and Charter has announced the nationwide launch of a new line of Xumo TVs from Westinghou...
15/07/2025
LONGMONT, Colo. DPA Microphones said Chris Kontopanos has joined the microphone solutions provider as its new regional sales manager, Mid-Atlantic....
15/07/2025
NOT FOR DISTRIBUTION IN OR INTO OR TO ANY PERSON LOCATED OR RESIDENT IN THE UNIT...
15/07/2025
News Corp Authorizes New $1 Billion Stock Repurchase Program and Intends to Acce...
15/07/2025
Back to All News
Netflix Announces the Start of Production of Los secretos de l...
15/07/2025
Harmonic's Hybrid, AI-Powered and Live Sports Streaming Innovations are Key ...
15/07/2025
The well-known and well-loved 4099 has been completely revamped. It features amazing CORE+ technology, which eliminates distortion and extends dynamic range, a ...
15/07/2025
Comscore Partners with HyphaMetrics to Launch Advanced Person-Level Audience Mea...
15/07/2025
Peak audiences of over half a million tune into both Kerry v Tyrone and Donegal ...
15/07/2025
PROTON Announces Production Facility Expansion Ahead of IBC2025 posted: 15/07/2025
Proton Announces Production Facility Expansion Ahead of IBC2025
Alongs...
15/07/2025
Submissions for NVIDIA's Plug and Play: Project G-Assist Plug-In Hackathon a...
14/07/2025
The Royal Moroccan Air Force hosted a C-130 award ceremony with L3Harris Technologies President of Intelligence, Surveillance and Reconnaissance Jason Lambert a...
14/07/2025
UK launch set for September, following on from recent US and Germany releases
London - UK, July 14, 2025 - Nielsen, a global leader in audience measurement, da...
14/07/2025
SHENZHEN, China Telycam has added control options for its portfolio of pan-tilt-zoom (PTZ) cameras with a new plug-in for Elgato's Stream Deck family of con...
14/07/2025
The awards are open to IBC2025 show exhibitors and offer a valuable platform for companies to raise awareness for the new products and solutions they will be la...
14/07/2025
Clear-Com is excited to announce a significant expansion of the award-winning Arcadia Central Station with new software updates. Together, these releases dram...
14/07/2025
Grass Valley and Daktronics today announced a strategic technology partnership that unites Grass Valley's live production expertise with Daktronics' lea...
14/07/2025
Cloud playout solutions provider, Veset, and swXtch.io, provider of cloud-based multicast and reliable networking solutions, are partnering to deliver multicast...
14/07/2025
Tintri , a DDN subsidiary and leader in AI-powered data management solutions, today announces a game-changing and award-winning refresh to its global Partner P...
14/07/2025
HighField AI, the broadcast industry's first agentic and multimodal AI platform for automating graphics production, is now commercially available. Developed...
14/07/2025
Encompass Digital Media has today announced that it has been delivering a complex global services project for DAZN, the world s leading sports entertainment pla...
14/07/2025
MASV (massive.io), the fastest, most reliable large file transfer platform trusted by over 10,000 global organizations, announced today it will showcase its nex...
14/07/2025
Calrec is introducing a series of usability, customisation and system enhancements across the entire range of Argo consoles at IBC 2025, on Stand 8.C47. Look ou...
14/07/2025
Fincons Group, IT business consultancy and systems integrator company with more than 40 years of experience in the market, is proud to announce a new engagement...
14/07/2025
Intinor returns to IBC to present the latest updates to its Direkt series. A key development on show is Intinor Direkt Management (IDM), a web-based interface f...
14/07/2025
Back to All News
Trailer Out Now! Netflix and YRF Entertainment's Mandala ...
14/07/2025
Back to All News
25 Former Athletes Seek Redemption in Final Draft,' Japan...
Fasten your seatbelts. NVIDIA Research is revving up a new deep learning engine that creates 3D object models from standard 2D images - and can bring iconic cars like the Knight Rider's AI-powered KITT to life - in NVIDIA Omniverse.
