We highlight Sentinel, a system for real-time in-situ intelligent video analytics on mobile computing platforms. Sentinel combines state-of-the-art techniques in HPC with Dynamic Mode Decomposition (DMD), a proven method for data reduction and analysis. By leveraging CUDA, our early system prototype achieves significantly better-than-real-time performance for DMD-based background/foreground separation on high-definition video streams, thereby establishing the efficacy of DMD as the foundation on which to build higher level real-time computer vision techniques. In this talk, we present an overview of the Sentinel system, including the application of DMD to background/foreground separation in video streams, and outline our current efforts to enhance and extend the prototype system.
We present a new method that accurately detects soft-body collisions, specifically the edge-edge collisions that most other methods would miss, interactively on modern GPUs. Our method guarantees that no pass-through will occur between objects by using interpolation equations to represent motion between time steps, yielding nearly exact collision times and responses. GPU acceleration via CUDA allows this method to operate at interactive rates.
We present the Visual Simulation Laboratory (VSL), an ongoing project aimed at bringing the power of GPU computing to a variety of DoD application domains. VSL is an open-source framework developed by the U.S. Army Research Laboratory and its collaborators designed to transform legacy workflows into immersive, end-to-end physics-based simulation and analysis tools. GPU computing facilitates combined simulation and visualization, enabling analysts to interact with a visual representation of not just the results, but of the computational mechanisms as well. This poster highlights VSL and demonstrates the potential of GPU computing to transform a variety of applications across the DoD.
The fast and accurate rendering of transparent objects is an open problem in computer graphics as it necessitates expensive fragment sorting on the GPU. We present initial findings in optimizing existing order-independent transparency (OIT) algorithms by reducing costly global thread synchronization. We leverage these improvements in a novel application of OIT to ballistic simulations used in vulnerability/lethality analysis software.
Prediction of radio frequency (RF) energy propagation in complex urban environments is of great interest when planning, optimizing and analyzing wireless networks. A tool for fast prediction could improve network coverage, provide estimates of signal strength, estimate time delay of multipath signals, and provide data for power allocation in the deployed transmitters. This poster highlights the Manta-RF system, which enables RF prediction through modifications to the Manta interactive ray tracing framework. We are actively migrating the current implementation to NVIDA's CUDA architecture. Ray tracing scales well with core count, and we expect to achieve near-interactive RF simulations as a result.
Explore recent advances in high-performance GPU ray tracing for applications other than optical rendering. In this session, we dive into the details of Rayforce, a CUDA ray tracing engine that leverages a new graph-based spatial indexing structure to achieve performance in excess of one billion rays per second in some non-trivial scenarios. We then explore several example applications that leverage Rayforce in a framework for cognition-driven simulation (CDS) that enables analysts to experience an immersive physics-based simulation environment. Compared to traditional means of analysis, CDS represents a next-generation approach to simulation and analysis across a broad range of application domains.