This site requires Javascript in order to view all its content. Please enable Javascript in order to access all the functionality of this web site. Here are the instructions how to enable JavaScript in your web browser.

NVIDIA at CVPR 2021

Connect with NVIDIA researchers at this year’s Computer Vision and Pattern Recognition (CVPR) online conference to learn more about the amazing work we're doing. Come see how NVIDIA Research collaborates with CVPR members to deliver AI breakthroughs across the community.

It all happens June 19–25, 2021.

Congratulations to the Contest Winners!

Rodolfo V.
Sophia A.
Andrew S.
Rahul D.

Niveditha K.
Andrew J.
Ishani V.
Grace L.

Thank you to all 706 participants who entered the drawing.

Mission AI Possible: NVIDIA Researchers Stealing the Show

Roll out of bed, fire up the laptop, turn on the webcam — and look picture-perfect in every video call, with the help of AI developed by NVIDIA researchers.

NVIDIA Vid2Vid Cameo, one of the deep learning models behind the NVIDIA Maxine SDK for video conferencing, uses generative adversarial networks (known as GANs) to synthesize realistic talking-head videos using a single 2D image of a person.

Watch Video

Trail Blazing our Future with AI and NVIDIA Research

In this chapter of I AM AI, we explore how NVIDIA Researchers across the world are bettering our world by doing their life’s work in AI research. Get a glimpse of the passion, dedication and curiosity that drives these brilliant minds and see the breadth of how AI is advancing humanity.

WATCH VIDEO

Schedule at a Glance

Presentations
Workshops and Tutorials

NVIDIA’s 28 accepted papers at this year’s online CVPR feature a range of groundbreaking research in the field of computer vision. From simulating dynamic gaming environments to powering coarse-to-fine neural architecture search for medical imaging, explore the work NVIDIA is bringing to the CVPR community.

Monday 6/21	Tuesday 6/22	Wednesday 6/23	Thursday 6/24	Friday 6/25
	Neural Parts: Learning Expressive Shape Abstractions with Invertible Neural Networks Despoina Paschalidou, Angelos Katharopoulos, Andreas Geiger, Sanja Fidler 6 - 8:30 a.m. EST \| Paper	DeepTag: An Unsupervised Deep Learning Method for Motion Tracking on Cardiac Tagging Magnetic Resonance Images Meng Ye, Mikael Kanski, Dong Yang, Qi Chang, Zhennan Yan, Qiaoying Huang, Leon Axel, Dimitris N. Metaxas 6 - 8:30 a.m. EST \| Paper	Neural Geometric Level of Detail: Real-Time Rendering with Implicit Shapes Towaki Takikawa, Joey Litalien, Kangxue Yin, Karsten Kreis, Charles T Loop, Alec Jacobson, Derek Nowrouzezahrai, Morgan McGuire, Sanja Fidler 6 - 8:30 a.m. EST \| Paper	See Through Gradients: Image Batch Recovery via GradInversion Hongxu Yin, Arun Mallya, Arash Vahdat, Jose M. Alvarez, Jan Kautz, Pavlo Molchanov 6 - 8:30 a.m. EST \| Paper
	Minimally Invasive Surgery for Sparse Neural Networks in Contrastive Manner Chong Yu 6 - 8:30 a.m. EST \| Paper	Self-Supervised Learning of Depth Inference for Multi-view Stereo Jiayu Yang, Jose M. Alvarez, Miaomiao Liu 6 - 8:30 a.m. EST \| Paper	Normalized Avatar Synthesis Using StyleGAN and Perceptual Refinement Huiwen Luo, Koki Nagano, Han-Wei Kung, Qingguo Xu, Zejian Wang, Lingyu Wei, Liwen Hu, Hao Li 6 - 8:30 a.m. EST \| Paper
		Self-Supervised Learning on Point Clouds by Learning Discrete Generative Models Benjamin Eckart, Wentao Yuan, Chao Liu, Jan Kautz 6 - 8:30 a.m. EST \| Paper	Optimal Quantization Using Scaled Codebook Despoina Paschalidou, Angelos Katharopoulos, Andreas Geiger, Sanja Fidler 6 - 8:30 a.m. EST \| Paper
		Semantic Segmentation with Generative Models: Semi-supervised Learning and Strong Out-of-Domain Generalization Daiqing Li, Junlin Yang, Karsten Kreis, Antonio Torralba, Sanja Fidler 6 - 8:30 a.m. EST \| Paper
Over-the-Air Adversarial Flickering Attacks against Recognition Networks Roi Pony, Itay Naeh, Shie Mannor 11a.m. - 1:30 p.m EST \| Paper	Towards Good Practices for Efficiently Annotating Large-Scale Image Classification Datasets Yuan-Hong Liao, Amlan Kar, Sanja Fidler 11 a.m. - 1:30 p.m. EST \| Paper	Learning Continuous Image Representation with Local Implicit Image Function Yinbo Chen, Sifei Liu, Xiaolong Wang 11a.m. - 1:30 p.m. EST \| Paper	Binary TTC: A Temporal Geofence for Autonomous Navigation Abhishek badki, Orazio Gallo, Jan Kautz, Pradeep Sen 11 a.m. - 1:30 p.m. EST \| Paper
	RGB-D Local Implicit Function for Depth Completion of Transparent Objects Luyang Zhu, Arsalan Mousavian, Yu Xiang, Hammad Mazhar, Jozef van Eenbergen, Shoubhik Debnath, Dieter Fox 11a.m. - 1:30 p.m.. EST \| Paper	Learning to Track Instances without Video Annotations Yang Fu, Sifei Liu, Umar Iqbal, Shalini De Mello, Humphrey Shi, Jan Kautz 11a.m. - 1:30 p.m. EST \| Paper
		Deep Two-View Structure-from-Motion Revisited Jianyuan Wang, Yiran Zhong, Yuchao Dai, Stan Birchfield, Kaihao Zhang, Nikolai Smolyanskiy, HONGDONG LI 11 a.m. - 1:30 p.m. EST \| Paper
		DexYCB: A Benchmark for Capturing Hand Grasping of Objects Yu-Wei Chao, Wei Yang, Yu Xiang, Pavlo Molchanov, Ankur Handa, Jonathan Tremblay, Yashraj S Narang, Karl Van Wyk, Umar Iqbal, Stan Birchfield, Jan Kautz, Dieter Fox 11 a.m. - 1:30 p.m. EST \| Paper
	DiNTS: Differentiable Neural Network Topology Search for Large-Scale Medical Image Segmentation Yufan He, Dong Yang, Holger R Roth, Can Zhao, Daguang Xu 10:00 p.m. – 12:30 a.m. EST \| Paper	Synthesizing Long-Term Human Motion and Interaction in Scenes Jiashun Wang, Huazhe Xu, Jingwei Xu, Sifei Liu, Xiaolong Wang 11 a.m. - 1:30 p.m. EST \| Paper
	Weakly Supervised Learning of Rigid Scene Flow Zan Gojcic, Or Litany, Andreas Wieser, Leonidas Guibas, Tolga Birdal 10 p.m. - 12:30 a.m. EST \| Paper	Weakly-Supervised Physically Unconstrained Gaze Estimation Rakshit S Kothari, Shalini De Mello, Umar Iqbal, Wonmin Byeon, Seonwook Park, Jan Kautz 10 p.m. - 12:30 a.m. EST \| Paper	IoUMatch: Leveraging IoU Prediction for Semi-Supervised Object Detection He Wang, Yezhen Cong, Or Litany, Yue Gao, Leonidas Guibas 10 p.m. - 12:30 a.m. EST \| Paper
	DriveGAN: Towards a Controllable High-Quality Neural Simulation Seung Wook Kim, Jonah Philion, Antonio Torralba, Sanja Fidler 10 p.m. - 12:30 a.m. EST \| Paper	One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing Ting-Chun Wang, Arun Mallya, Ming-Yu Liu 10 p.m. - 12:30 a.m. EST \| Paper	Semi-Supervised Hand-Object Poses Estimation with Interactions in Time Shaowei Liu, Hanwen Jiang, Jiarui Xu, Sifei Liu, Xiaolong Wang 10 p.m. - 12:30 a.m. EST \| Paper
	View Generalization for Single Image Textured Models Anand Bhattad, Aysegul Dundar, Guilin Liu, Andrew Tao, Bryan Catanzaro 10 p.m. - 12:30 a.m. EST \| Paper	DatasetGAN: Efficient Labeled Data Factory with Minimal Human Effort Yuxuan Zhang, Huan Ling, Jun Gao, Kangxue Yin, Jean-Francois V Lafleche, Adela Barriuso, Antonio Torralba, Sanja Fidler 10 p.m. - 12:30 a.m. EST \| Paper

Join NVIDIA at CVPR workshops and tutorials to engage with a diverse team of experts, expand your knowledge, and grow your skills with hands-on training.

Saturday 6/19	Monday 6/21	Tuesday 6/22	Wednesday 6/23	Thursday 6/24	Friday 6/25
LatinX in CV (LXCV) Research 8 a.m. - 7 p.m. EST \| Workshop					5th AI City Challenge Milind Naphade, Shuo Wang, Zheng Tang, David Anastasiu, Ming-Ching Chang, Pranamesh Chakraborty, Liang Zheng, Xiaodong Yang, Anuj Sharma, Stan Sclaroff, Rama Chellappa 8 a.m.- 2:00 pm. EST \| Workshop

Deep Dive

Fast Track to Production AI with NVIDIA Pre-trained Models and TLT

NVIDIA pre-trained models and Transfer Learning Toolkit (TLT) significantly reduce the time and costs associated with large-scale data collection and labeling, and eliminates the burden of training AI/ML models from scratch. The latest release includes new pre-trained models for pose estimation, people detection, and lots of computer vision use cases. It also supports training on AWS, GCP, and Azure cloud, as well as out-of-box deployment on NVIDIA Triton^™, DeepStream SDK, and Jarvis.

Learn More

NVIDIA Isaac Sim on Omniverse

Discover the only robotics simulation application and synthetic data generation tool that lets you develop, test, and manage robots in photorealistic, physically-accurate virtual worlds with unbeatable scalability. Next-gen NVIDIA Isaac Sim is built on NVIDIA Omniverse^™, accelerating robotics development with the powerful capabilities of the virtual simulation and collaboration platform.

Learn More

Omniverse Kaolin accelerates 3D deep learning research

Accelerate 3D Deep Learning Research with Kaolin and Omniverse Kaolin App

Omniverse Kaolin app is a powerful visualization tool that simplifies and accelerates 3D deep learning research using NVIDIA’s Kaolin PyTorch library. The Kaolin app leverages the Omniverse platform, USD format and RTX rendering to provide interactive tools that allow visualizing 3D outputs of any deep learning model as it is training, inspecting 3D datasets to find inconsistencies and gain intuition, and rendering large synthetic datasets from collections of 3D data. This lets you reduce the time needed to develop AI research for a wide range of 3D applications.

Watch the overview

NVIDIA Applied Research Acceleration Program

NVIDIA Applied Research Accelerator Program

This program supports research projects that have the potential to make a real-world impact through deployment into GPU-accelerated applications adopted by commercial and government organizations. It’s designed to accelerate development and adoption by providing access to technical guidance, hardware, and funding based on project requirements, maturity, and impact.

Learn More

DRIVE Developer Days now available On-Demand

The annual DRIVE Developer Days was held during GTC 2021, featuring a series of specialized sessions on autonomous vehicle hardware and software, including perception, mapping, simulation and more, all led by NVIDIA experts. These sessions are now available to view anytime with NVIDIA On-Demand.

Watch Now

NVIDIA Developer Program

Get the advanced tools and training you need to successfully build applications on all NVIDIA technology platforms.

Learn More

Featured Demos

Explore how NVIDIA technologies are transforming a variety of industries with powerful demos that highlight the latest breakthroughs in AI, data science, graphics, healthcare, and more.

GANverse3D: Knight Rider KITT Re-created with AI by NVIDIA

NVIDIA GANverse3D is a powerful breakthrough for 2D-to-3D models. Now, it's possible to create a 3D, animatable model with a single 2D picture.

Learn More

NVIDIA Maxine

NVIDIA Maxine^™ is a GPU-accelerated SDK with state-of-the-art AI features for developers to build virtual collaboration and content creation applications such as video conferencing and live streaming.

Learn More

NVIDIA Pre-Trained Models and TLT

Watch this body- pose estimation demo using the training support and NVIDIA Transfer Learning Toolkit from the latest release.

Watch Now

NVIDIA TAO, Metropolis, and Fleet Command

Discover a faster, easier, more effective way to adapt, deploy, and manage AI-enabled applications

Learn More

Flowers Classification Inference on NVIDIA A100 GPUs with MIG

NVIDIA A100 Multi-Instance GPU (MIG) technology allows multiple deep learning inference workload configurations.

Learn More

NVIDIA NGC Catalog

Collections from the NVIDIA NGC catalog are use-case-based content. They’re curated in easy-to-use packages that make it easy to discover the compatible framework containers, models, Jupyter notebooks, detailed documentation, and other resources to get started faster with your AI use cases.

Learn More

Factory of the Future in NVIDIA Omniverse

See how an entire factory, its virtual workers, and robots are simulated, operated, and maintained entirely in NVIDIA Omniverse with NVIDIA Isaac and Fleet Command.

Learn More

Watch an Autonomous Vehicle Navigate Stuttgart, Germany in High-Fidelity Simulation

The new NVIDIA DRIVE Sim^™ simulator uses the advanced Omniverse RTX Renderer’s real time ray tracing capabilities to generate accurate sensor data at real-time rates.

Watch Now

NVIDIA's Work in AI

Watch to see NVIDIA's AI research in action, including GANcraft Scene Generation, SyleGAN High-Res Image Generation, DLSS AI Rendering, Sim2Real Quadraped Locomotion, SimNet Physics-informed AI Model, and GANverse3D 2D to 3D Generation.

Watch Now

NVIDIA Jarvis

NVIDIA Jarvis is a GPU-accelerated framework for building multimodal conversational AI applications using state-of-the-art models that run in real-time.

Learn More

NVIDIA RESEARCH
AI PLAYGROUND

Discover our most recent AI research and the new capabilities deep learning brings to visual and audio applications. Explore the latest innovations and see how you can bring them into your own work. We'll update this page frequently with new demos and tools.

PLAY HERE

Inception Partners

LIKE NO PLACE
YOU’VE EVER WORKED

You’ll solve some of the world’s hardest problems and discover never-before-seen ways to improve the quality of life for people everywhere. From healthcare to robots, self-driving cars to blockbuster movies. And a growing list of new opportunities every single day. Explore all of our open roles, including internships and new college graduate positions.

Learn more about our career opportunities by exploring current job openings as well as university jobs.

Learn More

Monday 6/21	Tuesday 6/22	Wednesday 6/23	Thursday 6/24	Friday 6/25
	Neural Parts: Learning Expressive Shape Abstractions with Invertible Neural Networks Despoina Paschalidou, Angelos Katharopoulos, Andreas Geiger, Sanja Fidler 6 - 8:30 a.m. EST \| Paper	DeepTag: An Unsupervised Deep Learning Method for Motion Tracking on Cardiac Tagging Magnetic Resonance Images Meng Ye, Mikael Kanski, Dong Yang, Qi Chang, Zhennan Yan, Qiaoying Huang, Leon Axel, Dimitris N. Metaxas 6 - 8:30 a.m. EST \| Paper	Neural Geometric Level of Detail: Real-Time Rendering with Implicit Shapes Towaki Takikawa, Joey Litalien, Kangxue Yin, Karsten Kreis, Charles T Loop, Alec Jacobson, Derek Nowrouzezahrai, Morgan McGuire, Sanja Fidler 6 - 8:30 a.m. EST \| Paper	See Through Gradients: Image Batch Recovery via GradInversion Hongxu Yin, Arun Mallya, Arash Vahdat, Jose M. Alvarez, Jan Kautz, Pavlo Molchanov 6 - 8:30 a.m. EST \| Paper
	Minimally Invasive Surgery for Sparse Neural Networks in Contrastive Manner Chong Yu 6 - 8:30 a.m. EST \| Paper	Self-Supervised Learning of Depth Inference for Multi-view Stereo Jiayu Yang, Jose M. Alvarez, Miaomiao Liu 6 - 8:30 a.m. EST \| Paper	Normalized Avatar Synthesis Using StyleGAN and Perceptual Refinement Huiwen Luo, Koki Nagano, Han-Wei Kung, Qingguo Xu, Zejian Wang, Lingyu Wei, Liwen Hu, Hao Li 6 - 8:30 a.m. EST \| Paper
		Self-Supervised Learning on Point Clouds by Learning Discrete Generative Models Benjamin Eckart, Wentao Yuan, Chao Liu, Jan Kautz 6 - 8:30 a.m. EST \| Paper	Optimal Quantization Using Scaled Codebook Despoina Paschalidou, Angelos Katharopoulos, Andreas Geiger, Sanja Fidler 6 - 8:30 a.m. EST \| Paper
		Semantic Segmentation with Generative Models: Semi-supervised Learning and Strong Out-of-Domain Generalization Daiqing Li, Junlin Yang, Karsten Kreis, Antonio Torralba, Sanja Fidler 6 - 8:30 a.m. EST \| Paper
Over-the-Air Adversarial Flickering Attacks against Recognition Networks Roi Pony, Itay Naeh, Shie Mannor 11a.m. - 1:30 p.m EST \| Paper	Towards Good Practices for Efficiently Annotating Large-Scale Image Classification Datasets Yuan-Hong Liao, Amlan Kar, Sanja Fidler 11 a.m. - 1:30 p.m. EST \| Paper	Learning Continuous Image Representation with Local Implicit Image Function Yinbo Chen, Sifei Liu, Xiaolong Wang 11a.m. - 1:30 p.m. EST \| Paper	Binary TTC: A Temporal Geofence for Autonomous Navigation Abhishek badki, Orazio Gallo, Jan Kautz, Pradeep Sen 11 a.m. - 1:30 p.m. EST \| Paper
	RGB-D Local Implicit Function for Depth Completion of Transparent Objects Luyang Zhu, Arsalan Mousavian, Yu Xiang, Hammad Mazhar, Jozef van Eenbergen, Shoubhik Debnath, Dieter Fox 11a.m. - 1:30 p.m.. EST \| Paper	Learning to Track Instances without Video Annotations Yang Fu, Sifei Liu, Umar Iqbal, Shalini De Mello, Humphrey Shi, Jan Kautz 11a.m. - 1:30 p.m. EST \| Paper
		Deep Two-View Structure-from-Motion Revisited Jianyuan Wang, Yiran Zhong, Yuchao Dai, Stan Birchfield, Kaihao Zhang, Nikolai Smolyanskiy, HONGDONG LI 11 a.m. - 1:30 p.m. EST \| Paper
		DexYCB: A Benchmark for Capturing Hand Grasping of Objects Yu-Wei Chao, Wei Yang, Yu Xiang, Pavlo Molchanov, Ankur Handa, Jonathan Tremblay, Yashraj S Narang, Karl Van Wyk, Umar Iqbal, Stan Birchfield, Jan Kautz, Dieter Fox 11 a.m. - 1:30 p.m. EST \| Paper
	DiNTS: Differentiable Neural Network Topology Search for Large-Scale Medical Image Segmentation Yufan He, Dong Yang, Holger R Roth, Can Zhao, Daguang Xu 10:00 p.m. – 12:30 a.m. EST \| Paper	Synthesizing Long-Term Human Motion and Interaction in Scenes Jiashun Wang, Huazhe Xu, Jingwei Xu, Sifei Liu, Xiaolong Wang 11 a.m. - 1:30 p.m. EST \| Paper
	Weakly Supervised Learning of Rigid Scene Flow Zan Gojcic, Or Litany, Andreas Wieser, Leonidas Guibas, Tolga Birdal 10 p.m. - 12:30 a.m. EST \| Paper	Weakly-Supervised Physically Unconstrained Gaze Estimation Rakshit S Kothari, Shalini De Mello, Umar Iqbal, Wonmin Byeon, Seonwook Park, Jan Kautz 10 p.m. - 12:30 a.m. EST \| Paper	IoUMatch: Leveraging IoU Prediction for Semi-Supervised Object Detection He Wang, Yezhen Cong, Or Litany, Yue Gao, Leonidas Guibas 10 p.m. - 12:30 a.m. EST \| Paper
	DriveGAN: Towards a Controllable High-Quality Neural Simulation Seung Wook Kim, Jonah Philion, Antonio Torralba, Sanja Fidler 10 p.m. - 12:30 a.m. EST \| Paper	One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing Ting-Chun Wang, Arun Mallya, Ming-Yu Liu 10 p.m. - 12:30 a.m. EST \| Paper	Semi-Supervised Hand-Object Poses Estimation with Interactions in Time Shaowei Liu, Hanwen Jiang, Jiarui Xu, Sifei Liu, Xiaolong Wang 10 p.m. - 12:30 a.m. EST \| Paper
	View Generalization for Single Image Textured Models Anand Bhattad, Aysegul Dundar, Guilin Liu, Andrew Tao, Bryan Catanzaro 10 p.m. - 12:30 a.m. EST \| Paper	DatasetGAN: Efficient Labeled Data Factory with Minimal Human Effort Yuxuan Zhang, Huan Ling, Jun Gao, Kangxue Yin, Jean-Francois V Lafleche, Adela Barriuso, Antonio Torralba, Sanja Fidler 10 p.m. - 12:30 a.m. EST \| Paper

NVIDIA at CVPR 2021

Congratulations to the Contest Winners!

Mission AI Possible: NVIDIA Researchers Stealing the Show

Trail Blazing our Future with AI and NVIDIA Research

Schedule at a Glance

Deep Dive

Fast Track to Production AI with NVIDIA Pre-trained Models and TLT

NVIDIA Isaac Sim on Omniverse

Accelerate 3D Deep Learning Research with Kaolin and Omniverse Kaolin App

NVIDIA Applied Research Accelerator Program

DRIVE Developer Days now available On-Demand

NVIDIA Developer Program

Featured Demos

GANverse3D: Knight Rider KITT Re-created with AI by NVIDIA​

NVIDIA Maxine​

NVIDIA Pre-Trained Models and TLT

NVIDIA TAO, Metropolis, and Fleet Command

Flowers Classification Inference on NVIDIA A100 GPUs with MIG

NVIDIA NGC Catalog

Factory of the Future in NVIDIA Omniverse

Watch an Autonomous Vehicle Navigate Stuttgart, Germany in High-Fidelity Simulation

NVIDIA's Work in AI

NVIDIA Jarvis​

NVIDIA RESEARCH AI PLAYGROUND

Inception Partners

LIKE NO PLACEYOU’VE EVER WORKED

Sign up to receive the latest news from NVIDIA.

GANverse3D: Knight Rider KITT Re-created with AI by NVIDIA

NVIDIA Maxine

NVIDIA Jarvis

NVIDIA RESEARCH
AI PLAYGROUND

LIKE NO PLACE
YOU’VE EVER WORKED