Gowtham Elan’s Post

A Rerun Viewer for the DROID Dataset! DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset is a robot manipulation dataset by Alexander Khazatsky et al. with 76k demonstration trajectories or 350h of interaction data, collected across 564 scenes and 86 tasks. It is collected using a setup consisting of a Franka Panda 7DoF robot arm, two adjustable Zed 2 stereo cameras, a wrist-mounted Zed Mini stereo camera, and an Oculus Quest 2 headset with controllers for teleoperation (not included in the example). In the Rerun Viewer, you see a URDF of the robot arm, 2D and 3D from the Zed stereo cameras, and 2D from the wrist-mounted Zed Mini stereo camera. In addition, you get time-series for torques and velocities. You can find the code for the Viewer in the link in the comments!

Exploring the future of construction! Spent a chunk of today diving into how #TeamSkender is leveraging the latest tech and AI to redefine what’s possible in construction. It’s incredible to see the R&D behind selecting tech that truly adds value for our clients and partners (h/t Ben Stocker). I was especially excited to read this NVIDIA blog that showcases two Skender projects! With the power of AI, RTX GPUs, photogrammetry, LiDAR and gaussian splats, builders like Skender are capturing, modeling and visualizing projects with precision like never before. These #contech innovations are just a few of the tools that allow us to gather data faster and smarter, delivering even better insights, efficiency and quality to our clients. Next-level, inspiring stuff! Read NVIDIA's article: https://lnkd.in/gb5q9wuv

Playing around with a new interactive environment for the Move Real-Time system. This is markerless motion capture powered by AI and physics based models. This example used 7 FLIR cameras with an NVIDIA A6000 and a Live Link Plugin to Unreal Engine. #moveai #motioncapture #3d #3danimation #unrealengine

🎅 Nikolaustag in Simspector! 🎅 Today’s episode of our Sneak Peek Fridays comes with a festive twist! It seems Nikolaus made his way into Simspector, our perception sensor analysis tool! 🎄✨ Using 3D geometry and material assignment analysis capabilities, we spotted him on a German highway today in stunning detail—complete with a sleigh and all the holiday cheer! 🎁 This feature will also cover the upcoming ASAM OpenMATERIAL standard, showcasing Simspector’s adaptability and precision. 🚀 Apparently, his sleigh is guided by a Lidar system—no wonder he never misses a rooftop! 🎅✨ This discovery highlights not just the festive spirit but also the growing role of advanced sensor technology in guiding even the most magical journeys. Check out the video below to see Nikolaus in action. While it’s a bit of festive fun, it also highlights the powerful tools we’re developing to support credible perception simulation! 🌟 #Nikolaustag #Simspector #Perception# Sensor #Simulation #OpenMATERIAL #CredibleSimulation

Announced at #GTC24, NVIDIA Isaac Perceptor, optimized on Jetson Orin, uses multiple cameras for 3D surround perception to detect hazards invisible to standard 2D lidar. Using robust AI-based depth estimation, GPU-accelerated 3D reconstruction, and semantic segmentation, see how a mobile robot can work more safely alongside humans.

Announced at #GTC24, NVIDIA Isaac Perceptor, optimized on Jetson Orin, uses multiple cameras for 3D surround perception to detect hazards invisible to standard 2D lidar. Using robust AI-based depth estimation, GPU-accelerated 3D reconstruction, and semantic segmentation, see how a mobile robot can work more safely alongside humans.

Announced at #GTC24, NVIDIA Isaac Perceptor, optimized on Jetson Orin, uses multiple cameras for 3D surround perception to detect hazards invisible to standard 2D lidar. Using robust AI-based depth estimation, GPU-accelerated 3D reconstruction, and semantic segmentation, see how a mobile robot can work more safely alongside humans.