Loading...

Sessions

  • Other Sessions
    • Intermediate Technical
      Startups/NVIDIA Inception
      Talks & Panels
      RTX GPU
      CUDA
      TensorRT
      Research: Non-Academic
      Applied AI
      All
      8:00 a.m. Thursday, Mar 23
      Thursday, Mar 23
      Neural avatar solutions currently need to overcome three challenges: reliability at scale, minimal or no offline data capture complexity, and minimal training/inference complexity. We present some avenues to resolve these by merging neural avatar proposals together with conventional video/3D encoding standards. Such a merger can bridge the gap between traditional "codecs" and photorealistic neural avatars and offer significant runtime and bit-rate efficiency versus all existing work. Importantly, it can also anchor the rendered output to the physical reality captured by sensors at the sender side, which ensures any scene/person augmentation or reenactment is interpretable and controllable when deployed at scale. We'll show demonstration results on NVIDIA RTX GPUs and summarize lessons learned, such as the importance of accurate visual quality scoring.

      , CTO and Professor, University College London, iSIZE

      • Thursday, Mar 233:00 PM - 3:30 PM GMT
    • 14
      3
      3
      3
      Intermediate Technical
      Talks & Panels
      RTX GPU
      DXR
      Developer / Engineer
      Graphics & Media
      All
      Wednesday, Mar 22
      3:00 a.m. Wednesday, Mar 22
      Raytracing algorithms rely on having a spatial acceleration structure built over the scene geometry, which scales poorly as the geometric complexity grows. Micro-meshes are designed to address this important problem. They are central to a powerful framework that enables efficient handling of highly detailed geometry in modern rendering pipelines. At its core, it leverages a representation that explicitly extracts high-frequency geometric detail and encodes it in a set of displacement or opacity maps, specified relative to a lower resolution base mesh. This decomposition is key to both its efficiency and the scalability of the representation: first, the encoding of these maps can employ an effective compression scheme, thanks to their regular, and explicitly hierarchical structure; second, the lower-resolution geometry, represented as a triangular mesh, can be used directly in rendering algorithms that operate on coarse-grained geometry. As a result, raytracing acceleration structures can use this coarse-grained scene representation to achieve much better performance scaling. In this session, we’ll provide an overview of key ideas and concepts behind micro-meshes, and demonstrate their effectiveness in representing high-resolution geometry assets, both for raytracing and rasterization pipelines. Attendees will learn how micro-mesh technology can be integrated in their content creation flows and rendering engines, leveraging the latest hardware, tools and SDKs provided by NVIDIA.

      , Senior Distinguished Engineer, NVIDIA

      , Developer Technology Engineer, NVIDIA

      • Wednesday, Mar 2210:00 AM - 10:55 AM GMT
    • 5
      Intermediate Technical
      Talks & Panels
      Maxine
      Developer / Engineer
      Applied AI
      All
      10:00 a.m. Thursday, Mar 23
      Thursday, Mar 23
      Live communications services and platforms are fundamental to how we interact with each other, and they're growing at exponential rates, facing quality challenges that severely impact understanding. Whether for a video conference, a call made to a customer service center, or a livestream, Maxine enables clear communications to enhance virtual interactions with purposely built real-time AI technologies. Increasing wireless connectivity around the world enables people to dial in from far more locations than before. Quality of audio and video becomes critically important when the caller is on the move or in locations with poor connectivity. NVIDIA Maxine is a suite of GPU-accelerated SDKs and cloud-native microservices for deploying AI features that enhance audio, video, and augmented reality effects in real time. Maxine’s state-of-the-art models create high-quality effects that can be achieved even with standard microphones and webcams. We'll walk you through Maxine's latest features, including: • Updated improvements in Maxine SDKs, like eye-contact and audio super resolution; • New standalone deployable microservices, like RIVA speech-to-speech integrated in video conferencing pipeline with other Maxine effects for fluent interactions in your native language; and • Applications that supercharge your preferred video conferencing tool with Maxine capabilities.

      , Professional Broadcast Product Manager, NVIDIA

      , Senior Manager, System Software, NVIDIA

      , System Software Manager, NVIDIA

      • Thursday, Mar 235:00 PM - 5:30 PM GMT
    • Beginner Technical
      Special Event
      Riva
      Business Executive
      Applied AI
      All
      Monday, Mar 20
      8:00 a.m. Monday, Mar 20
      To maintain world-class platforms that meet customer expectations across the globe, progressive companies are adopting highly accurate and performant real-time speech AI – localized to support required languages -- and integrating it within the existing cloud, hybrid, or on-premises systems and applications. We'll see how Avaya is leveraging the capabilities of Riva speech AI technology to transform employee and customer experiences. We will also look at exciting future capabilities being explored "behind the curtain!" This talk is a must-attend for businesses looking to maximize the potential of AI Speech technology to accelerate their business goals.

      , Avaya Marketing DIrector, Avaya

      • Monday, Mar 203:00 PM - 3:30 PM GMT