Two NVIDIA Research papers - one exploring diffusion-based generative AI models and another on training generalist AI agents - have been honored with NeurIPS 2022 Awards for their contributions to the field of AI and machine learning.
These are among more than 60+ talks, posters and workshops with NVIDIA authors being presented at the NeurIPs conference, taking place this week in New Orleans and next week online.
Synthetic data generation - for images, text or video - is a key theme across several of the NVIDIA-authored papers. Other topics include reinforcement learning, data collection and augmentation, weather models and federated learning.
AI is an incredibly important technology, and NVIDIA is making fast progress across the gamut - from generative AI to autonomous AI agents, said Jan Kautz, vice president of learning and perception research at NVIDIA. In generative AI, we are not only advancing our theoretical understanding of the underlying models, but are also making practical contributions that will reduce the effort of creating realistic virtual worlds and simulations.
Reimagining the Design of Diffusion-Based Generative Models Diffusion-based models have emerged as a groundbreaking technique for generative AI. NVIDIA researchers won an Outstanding Main Track Paper award for work that analyzes the design of diffusion models, proposing improvements that can dramatically improve the efficiency and quality of these models.
The paper breaks down the components of a diffusion model into a modular design, helping developers identify processes that can be adjusted to improve the performance of the entire model. The researchers show that their modifications enable record scores on a metric that assesses the quality of AI-generated images.
Training Generalist AI Agents in a Minecraft-Based Simulation Suite While researchers have long trained autonomous AI agents in video-game environments such as Starcraft, Dota and Go, these agents are usually specialists in only a few tasks. So NVIDIA researchers turned to Minecraft, the world's most popular game, to develop a scalable training framework for a generalist agent - one that can successfully execute a wide variety of open-ended tasks.
Dubbed MineDojo, the framework enables an AI agent to learn Minecraft's flexible gameplay using a massive online database of more than 7,000 wiki pages, millions of Reddit threads and 300,000 hours of recorded gameplay (shown in image at top). The project won an Outstanding Datasets and Benchmarks Paper Award from the NeurIPS committee.
As a proof of concept, the researchers behind MineDojo created a large-scale foundation model, called MineCLIP, that learned to associate YouTube footage of Minecraft gameplay with the video's transcript, in which the player typically narrates the onscreen action. Using MineCLIP, the team was able to train a reinforcement learning agent capable of performing several tasks in Minecraft without human intervention.
Creating Complex 3D Shapes to Populate Virtual Worlds Also at NeurIPS is GET3D, a generative AI model that instantly synthesizes 3D shapes based on the category of 2D images it's trained on, such as buildings, cars or animals. The AI-generated objects have high-fidelity textures and complex geometric details - and are created in a triangle mesh format used in popular graphics software applications. This makes it easy for users to import the shapes into 3D renderers and game engines for further editing.
Named for its ability to Generate Explicit Textured 3D meshes, GET3D was trained on NVIDIA A100 Tensor Core GPUs using around 1 million 2D images of 3D shapes captured from different camera angles. The model can generate around 20 objects a second when running inference on a single NVIDIA GPU.
The AI-generated objects could be used to populate 3D representations of buildings, outdoor spaces or entire cities - digital spaces designed for industries such as gaming, robotics, architecture and social media.
Improving Inverse Rendering Pipelines With Control Over Materials, Lighting At the most recent CVPR conference, held in New Orleans in June, NVIDIA Research introduced 3D MoMa, an inverse rendering method that enables developers to create 3D objects composed of three distinct parts: a 3D mesh model, materials overlaid on the model, and lighting.
The team has since achieved significant advancements in untangling materials and lighting from the 3D objects - which in turn improves creators' abilities to edit the AI-generated shapes by swapping materials or adjusting lighting as the object moves around a scene.
The work, which relies on a more realistic shading model that leverages NVIDIA RTX GPU-accelerated ray tracing, is being presented as a poster at NeurIPS.
Enhancing Factual Accuracy of Language Models' Generated Text Another accepted paper at NeurIPS examines a key challenge with pretrained language models: the factual accuracy of AI-generated text.
Language models trained for open-ended text generation often come up with text that includes nonfactual information, since the AI is simply making correlations between words to predict what comes next in a sentence. In the paper, NVIDIA researchers propose techniques to address this limitation, which is necessary before such models can be deployed for real-world applications.
The researchers built the first automatic benchmark to measure the factual accuracy of language models for open-ended text generation, and found that bigger language models with billions of parameters were more factual than smaller ones. The team proposed a new technique, factuality-enhanced training, along with a novel sampling algorithm that together help train language models to generate accurate text - and demonstrated a reduction in the rate of factual errors from 33% to around 15%.
There are more th
Most recent headlines
17/02/2025
TVBEurope talks to the team who brought a piece of broadcasting history back to life on BBC Wales last night
By Jenny Priestley
Published: February 14, 2025 ...
17/02/2025
Eligible studios will receive the relief directly from their local authorities, with no need to submit a claim
By Matthew Corrigan
Published: February 17, 20...
17/02/2025
As DEI initiatives come under pressure in the USA, supporters of Rise, which promotes diversity in the media and entertainment sector, discuss the continuing im...
17/02/2025
EIB Global loan will partially fund SES's continued O3b mPOWER medium earth orbit (MEO) fleet expansion enabling additional capacity worldwide. O3b mPOWER s...
15/02/2025
At Nielsen, Business Resource Groups (BRGs) play a critical role in nurturing a ...
15/02/2025
Innovation to empower creativity: Explore the latest in Lightroom
Rob Christensen February 15, 2025
0 Comments
Whether you want to remove distracting ...
15/02/2025
Wisycom, a worldwide leader in RF technology and provider of wireless solutions, announces a new distribution agreement with EQUAPHON. An Argentinian profession...
15/02/2025
DPA Microphones, the leading manufacturer of high-quality microphone solutions, is pleased to welcome Nicholas Nick Mariano and Vincent Vince Divine to the ...
15/02/2025
Tuxera, a leading provider of quality-assured file systems and networking solutions, will showcase how Tuxera Fusion SMB enables unprecedented performance for m...
15/02/2025
Bitmovin's latest Video Developer Report indicates an increasing role of AI, the slow adoption of next-generation codecs, and continuing challenges when it ...
15/02/2025
When Michelin Enterprises set out to promote its prestigious Hotel Guide in the new Sleepless in the Kitchen , campaign, veteran German/Slovenian gaffer Robert...
15/02/2025
Formula E, the world's first all-electric racing series, revolutionizing motorsport with cutting-edge technology and a commitment to sustainability has sele...
15/02/2025
Agile Content, a global leader in TV and video technology solutions, announces its rebranding to AgileTV.
The transition to AgileTV stems from a commitment to ...
15/02/2025
Hiltron Communications will promote the latest additions to its range of satellite communication products, systems and supporting services at GovSatCom 2025. Th...
15/02/2025
Following the success of the original Airglow Booklight, DoPchoice debuts a new addition to their inflatable line the Float. This lightweight, inflatable, round...
15/02/2025
LAS VEGAS Marshall Electronics has announced that it will debut its newest camera controller, the RCP Plus, at the 2025 NAB Show at Booth N2649 between April 5 ...
15/02/2025
ATLANTA Gray Media has named Blake Sebo as the next general manager of KTUU and KYES, Gray's NBC and CBS affiliates in Anchorage, Alaska....
15/02/2025
KOKKEDAL, Denmark DPA Microphones will feature its new CORE+ technology for distortion-free microphone sound at the 2025 NAB Show, April 5-9, in Las Vegas....
15/02/2025
SWANSEA, U.K. QuickLink will launch StudioPro-NDI and StudioPro-3, two new models of its QuickLine StudioPro video production platform, at the 2025 NAB Show, Ap...
15/02/2025
Live From 2025 NBA All-Star Game Weekend: After 25 Years, Turner Sports All-In F...
15/02/2025
NASCAR Driver Cam on Max Embraces Cloud-Based Workflow for 2025 Season Chris Brown of WBD Sports says Mobii will play key part in delivering signals from the tr...
14/02/2025
Last month, we officially launched the Spotify Partner Program, a powerful new m...
14/02/2025
From an island off the coast of Lutruwita to the global stage in France SBS & ...
14/02/2025
The future of electro-optical/infrared (EO/IR) land systems is evolving so that ...
14/02/2025
Here are 5 reasons why you should attend ISE 2025
Discover more about Calrec: Learn more about Calrec's legacy of innovation in broadcast audio solutions ...
14/02/2025
World Radio Day: On the same wavelength
By Graham Murray, International Sales Manager at Calrec
To mark World Radio Day, Calrec's International Sales Ma...
14/02/2025
DALLAS and TYSONS, Va. Tegna has concluded a broadcast rights agreement that makes Tegna's KFAA in Dallas the exclusive local television partner of the Dall...
14/02/2025
Amazon's Prime Video has released a new app for Apple TV that the streamer said delivers an improved streaming experiences with faster scrolling, improved s...
14/02/2025
CLEVELAND Telos Alliance announced a software update for its Axia Quasar XR and SR line of audio-over-IP (AoIP) mixing consoles and will show the offering at th...
14/02/2025
Puget Systems Returns to HPA Tech Retreat 2025 with New Analysis Tools for Puge...
14/02/2025
Glass Animals: Tour of Earth Uses Blackmagic Live Production Workflow
Brie Clayton February 14, 2025
0 Comments
80six deploys Blackmagic URSA Broadcas...
14/02/2025
A dispute between Paramount and Youtube TV that would have pulled CBS stations and Paramount's cable channels from the streaming service has been temporaril...
14/02/2025
Colourist Deidre (Dee) McClelland details how she helped shape the distinct look and feel of whimsical Oscar-nominated stop-motion film, Memoir of a Snail
By C...
14/02/2025
BitFire said it intends to introduce forward-looking services that simplify the launch of live production workflows, and enhance real-time collaboration
By Jen...
14/02/2025
The claim by the liquidators is the latest twist in the ongoing saga of the companys collapse at the end of 2021
By Matthew Corrigan
Published: February 14, ...
14/02/2025
A carriage dispute between YouTube TV and Paramount could come to a head today (Feb. 13), with the virtual pay TV provider threatening to pull all of the progra...
14/02/2025
ENGLEWOOD CLIFFS, N.J. LG Electronics USA has opened a new in-house, state-of-the-art virtual production studio at its 360,000-square-foot North American headqu...
14/02/2025
NEW YORK In a bid to help local broadcasters find new ways of monetizing their content, the Coalition for Innovative Media Measurement (CIMM) and TVB have relea...
14/02/2025
WASHINGTON Three Democratic U.S. Senators have sent a stern letter blasting FCC Chair Brendan Carr for the weaponization of the agency for attacking broadcast...
14/02/2025
Years after launching the streaming service Apple TV+ in November of 2019, Apple has finally released a version of its steaming app for Android devices....
14/02/2025
FOOTHILL RANCH, Calif. Red Digital Cinema will make its V-RAPTOR [X] and KOMODO-X camera systems available with Nikon's premier Z Mount, providing filmmaker...
14/02/2025
DAYTONA, Fla. The Fox Sports production of this year's Daytona 500 will see the addition of a few new tech elements and the continued unfolding of the broad...
14/02/2025
Watch Bryson Battles Blind Audition on The Voice Get a Four-Chair Turn Michael Bubl said the Boston Conservatory at Berklee musical theater senior might have...
14/02/2025
2025 Daytona 500: FOX Sports Looks To Ride Super Bowl LIX Momentum Into NASCAR S...
14/02/2025
14 Feb 2025
VEON's Kyivstar and the Come Back Alive Foundation Raise USD 2....
14/02/2025
Max Renews THE PITT, Starring Noah Wyle, For A Second Season
Max
Warner Bros. Television Group
The Pitt
New Episodes Of The Medical Drama From John...
14/02/2025
This year, Technicolor proudly celebrates 110 years of shaping the magic of cinema pioneering color, visual effects, and innovation that bring unforgettable l...
14/02/2025
NBA All-Star: More Games, More Teams, More Music, More Sound From the Court Turner Sports' team will deploy familiar tech to capture all the audio By Dan D...
14/02/2025
LIV Golf Continues To Innovate With Drones, Data-Driven Graphics, and All-Access...
Two NVIDIA Research papers - one exploring diffusion-based generative AI models and another on training generalist AI agents - have been honored with NeurIPS 2022 Awards for their contributions to the field of AI and machine learning.
