Explore featured sessions for robotics, commercial applications, robotics simulation, reinforcement learning, robotics research, and robotics education. Also, learn how to get started on NVIDIA Jetson™.

Find out more about the latest breakthroughs in AI and robotics, simulation and training, robotics research and reinforcement learning, and automation in manufacturing.

Featured Speakers

Jens Lambrecht
Brian Gerkey

Open-Source Robotics at 20

Brian Gerkey
Co-founder and CEO, Open Robotics

Rafaello Bonghi
Hammad Mahzar

Sessions By Topics

Computer Vision - IVA/AI Cities

  • Accelerating the Development of Next-Generation AI Applications with DeepStream 6.0

    • Alvin Clark, NVIDIA
    • Carlos Garcia-Sierra, NVIDIA

    Edge AI and distributed processing applications are at the forefront of the deep learning revolution. Learn how DeepStream can help you create the next big thing for retail, manufacturing, health care, smart cities, and beyond. DeepStream 6.0 delivers , for the first time, temporal video processing that unlocks new use cases that range from action recognition to signal processing. Using the power of DeepStream has never been easier, with the introduction of Graph Composer, the new interface that lets developers create complex applications using nothing more than drag-and-drop operations. These applications can then be deployed as a container to any NVIDIA GPU-accelerated cloud or edge node.

  • Quickly Develop Vision AI Applications from Any Perspective

    • Chintan, Sr. Product Manager, NVIDIA
    • Akhil Docca, Sr. Product Manager, NVIDIA

    Today’s businesses are relying on computer vision to improve product quality, enable autonomous shopping, detect early onset of diseases, and more. Bringing production-ready Vision AI applications to market without massive training datasets and an army of data scientists is a daunting task for many. By applying transfer learning on pre-trained models, developers can build their AI models faster with minimal coding. We’ will show how to leverage the TAO Toolkit and pre-trained models with help of customer examples, demos, and sample applications to get you up and running in minutes.

  • The Rise of Smart Infrastructure: Automating Smart Spaces with NVIDIA Metropolis and Edge AI

    From retail shops to warehouses to our city streets, every space is becoming smarter thanks to AI-enabled vision applications. In this talk, NVIDIA will share how the world’s smartest spaces are leveraging NVIDIA Metropolis AI application framework to accelerate time to impact — from development to deployment and management. We will cover application areas; the breadth of NVIDIA tools for training, inference, optimization; and dev-ops management. We’ll also share how NVIDIA Metropolis simplifies solution integration by bringing together an entire rich ecosystem — from application developers, hardware partners, and system integrators.

  • How to Quickly Develop and Stand Up Smart Infrastructure Solutions with AI Launchpad and Metropolis

    Developing and deploying smart applications across a broad set of industries is now simpler than ever. By combining cloud native development, fully managed edge AI deployments and instant access to EGX cloud instances, you can now spend your time focused on application development rather than the worries of hardware availability or complex IT infrastructure deployments. In this session, you will learn how to develop Metropolis applications that can be deployed via Fleet Command out to globally distributed AI Infrastructure provided by AI LaunchPad.

Conversational AI

  • Your First Steps to Designing an Intelligent Assistant for Hands-free Applications

    Practical conversational AI systems like intelligent assistants (IA) now provide spoken responses to users’ requests in a fraction of a second, thanks to accelerated computing. Learn the initial steps to create an example IA for a hands-free application-- performing an upgrade to an aircraft while keeping maintenance notes. We examine the enablers for an IA based on NVIDIA Riva SDK, including a consideration of design trade-offs for the STT subsystem, options for question answering (QA), the application of neural machine translation in QA, and how to provide verbal responses in high-quality synthetic speech. We further examine the role of accelerated computer vision for an intelligent assistant to guide technicians in their work in real time, all while monitoring visually to prevent errors. Finally, we demonstrate programming with Riva, highlighting the APIs you need to call on your path to creating your own conversational AI system.

  • Conversational AI Demystified, and a Hands-On Walkthrough

    • Miguel Martínez, NVIDIA and SVA

    Conversational AI technologies are becoming ubiquitous, with countless products taking advantage of automatic speech recognition, natural language understanding, and speech synthesis coming to market. Thanks to new tools and technologies, developing conversational AI applications is easier than ever, enabling a much broader range of applications, such as virtual assistants, real-time transcription, and many more. The talk will start with an overview of the conversational AI landscape and discuss how any organization can get started developing conversational AI applications today. Then, we'll demonstrate how to build and deploy end-to-end conversational AI pipelines using NVIDIA RIVA.

Breakthroughs in Reinforcement Learning

  • Improve Agents without Retraining: Parallel Tree Search with Off-Ppolicy Correction

    • Shie Mannor, NVIDIA

    Tree Search (TS) is crucial to some of the most influential successes in reinforcement learning. We'll tackle two major challenges with TS that limit its usability-distribution shift and scalability. We’ll report a counter-intuitive phenomenon: action selection through TS and a pre-trained value function often leads to lower performance compared to the original pre-trained agent, even when having access to the exact state and reward in future steps. We'll show this is due to a distribution shift to areas where value estimates are highly inaccurate, and explain how to correct it. We'll then address the scalability issue given by the computational complexity of exhaustive TS that scales exponentially with the tree depth, and introduce a GPU breadth-first search that advances all nodes in each depth of the tree, simultaneously reducing runtime by two orders of magnitude.

  • Isaac Gym and Omniverse: High- Performance Reinforcement Learning Evolved

    • Gavriel State, NVIDIA
    • Lukasz Wawrzyniak

    We'll provide the latest updates on NVIDIA's Isaac Gym environment for high-performance reinforcement learning on GPU, including both the standalone version and upcoming work that fully integrates tensor-API accelerated physics simulation with the Omniverse platform. We'll review how the Isaac Gym system allows developers to create tensor-based views of physics states for all environments, explore key new API features, and demonstrate new training environments for physics-based animation, operational space control, and dexterous manipulation with Sim2Real transfer of learned policies to physical robots.

See autonomous machines session highlights from the previous GTC. Get ready for what’s to come.

Find the complete GTC On-Demand playlist here.

Explore All Session Topics

NVIDIA Developer Program

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

Accelerate your Startup

Explore the startup track at GTC to learn how NVIDIA Inception can fuel your growth through go-to-market support, world-class training, and technology assistance.

Get Hands-On Training

Interested in developing key skills in AI, accelerated data science, or accelerated computing? Get hands-on instructor-led training from the NVIDIA Deep Learning Institute (DLI) and earn a certificate demonstrating subject matter competency.