Seeed Studio’s Post

Insight review from #GTC2024💫 It's an honor to connect with Doruk Sönmez, the Intelligent Video Analytics Engineer from OpenZeka, at the NVIDIA Robotics GTC booth. We're diving deep into the transformative power of video analytics and LLM right at the edge. This demo showcased the exceptional capabilities of the Cordatus AI platform, enhanced by the potent inferencing capabilities of the reComputer Industrial Jetson Orin edge device. Witness firsthand how we're harnessing raw camera streams to deliver actionable, real-time insights, optimizing security and operational efficiency across industries. Check out reComputer Industrial Jetson Orin NX edge device: https://lnkd.in/gDpq3CT4 #nvidia #jetson #videoanalytics #LLM #EdgeAI #security

#GTC2024 😎 moments review: Join Roboflow and explore how to seamlessly deploy YOLO-World zero-shot object detection model on #reComputer Industrial edge device powered by NVIDIA Robotics Jetson Orin Nano! The whole pipeline transcends traditional boundaries with its open vocabulary capability, allowing it to recognize objects beyond predefined categories—making it more efficient and adaptable for real-world applications. Experience the flexibility of dynamically adjusting detection vocabularies to meet diverse needs without compromising performance. 👉 To get hands on the YOLO-World approach on Jetson Orin Nano 8GB, you may want to check out our reComputer Industrial J3011: https://lnkd.in/gcfq6g8D 🏂 YOLO-World GitHub repo: https://lnkd.in/gvASzQUX #nvidia #jetson #computervision #edgeai #zeroshot #objectdetection #yoloworld

#ProjectSpotlight Create a head-mounted AI navigation and detection system with NVIDIA Robotics Jetson Nano for the virtual impaired in a complex environment. Check out all components needed for this project: https://lnkd.in/g9wVk86M Key Features: ✅ Voice Activation: System operations begin with a simple voice command, processed through a TCP socket for seamless speech recognition. ✅ Environmental Perception: Leverage real-time object detection with MobileNet to map surroundings and calculate distances with image inputs. ✅ Tactile Guidance: Experience environmental awareness through a matrix of haptic vibrators. These vibrators vary in intensity based on object proximity, translating visual information into tactile feedback. Kickstart your vision AI project with the Jetson Nano edge device: https://lnkd.in/geJaQpYR #nvidia #jetsonNano #objectdetection #edgeAI #AIForGood #AccessibilityTech #Innovation

We can all agree : simulations can be fun when they go wrong ! A first interaction with NVIDIA's Isaac Sim, where I coordinated a handoff between an NVIDIA JETbot and Franka Emika robot🤖🤝 Accessing the remote RTX-powered system running the Omniverse App was a breeze using the Omniverse Streaming Client (Telnet protocol). This is just the beginning of a much larger project. 🌟 Project #goals: 1️⃣ Programmatically tap into Isaac Sim using the KIT interface. 2️⃣ Navigate robots with task logic in a simulation loop. 3️⃣ Develop scalable modules for repeated experiments. Introduction to Robotic Simulations in Isaac Sim by NVIDIA, #recommended #TechInnovation #Robotics #NVIDIA #IsaacSim #Omniverse #RoboticSimulations #FutureOfTech #Automation #AI #MachineLearning

InOrbit.AI was thrilled at the recent opportunity to #demo something new and very exciting at #GTC2024. We've been working hard on bridging the #Sim2Real gap, bringing #RobOps and advanced #simulations together, alongside fantastic collaborators at Ekumen and NVIDIA. Now, leveraging the NVIDIA Isaac Sim™ extensible #robotics simulator powered by NVIDIA Omniverse™, we've moved one step closer to closing the gap, with 'Real to Sim' and InOrbit Time Capsule.  InOrbit Time Capsule facilitates #incidentmanagement and robot fleet #optimization. It allows users to easily review robot #sensor data over time, track mission paths, and #analyze incidents with precision. Real to Sim is a 3D reconstruction of past events added to Time Capsule, where real-world data feeds advanced simulations for comprehensive analysis and optimization. Explore Real to Sim in the video below and learn more about InOrbit RobOps simulations at https://lnkd.in/gkpu-pnN NVIDIA GTC NVIDIA Robotics #GTC #GTC24 #techevent #conference #sim #sim2real #softwaredefined #simulation #simulationtechnology #omniverse #amr #robotics #Isaacsim #sim #demo #nvidiarobotics

Real to Sim with NVIDIA Isaac Sim and InOrbit.AI, powered by #Ekumen 📺 🤖 Check out this demo presented at #GTC2024 which showcases how #Ekumen can help you simulate your robots with high fidelity on their production environment, which is useful to reproduce issues and get to their root cause by leveraging #RobOps tools. Are you looking into growing your team with talented engineers capable of building simulations from the ground up, as well as maintaining and improving existing solutions? Contact our team and learn more about how we can collaborate to make it happen: 👉 contact@ekumenlabs.com #demo #robotics #robots #isaacsim #omniverse #nvidia #inorbit #simulation #simulator #software #engineeringservices #poweringyouringenuity

#computervision made easy! 🚀 To quickly determine which YOLO model best suits your application dataset in the field, try the Jetson-example. https://lnkd.in/gze_xs86 Firstly >> pip install jetson-examples After reboot >> reComputer run ultralytics-yolo We built this tool and made it the easiest way to test all #YOLOv8 models as well as your custom models!

🚀Finetune #LLM by Llama-Factory on Jetson Orin! Now you can tailor a custom private local LLM to meet your requirements. More info at Jetson examples GitHub 👉 https://lnkd.in/gkByK6Ym Quickly get started: ✅ Llama-Factory: an efficient tool to unify efficient Fine-Tuning of 100+ LLMs. ✅ Jetson-examples: a comprehensive toolkit for deploying containerized applications on NVIDIA Jetson devices. ✅ Workflow: use the alpaca_zh dataset to fine-tune the Phi-1.5 model, enabling it to have Chinese conversational capabilities. Discover reComputer J4012 Edge device powered by NVIDIA Robotics Jetson Orin NX 16GB to build your own LLM system: https://lnkd.in/g7J7w7Ts #NVIDIA #jetson #generativeai #llm #finetune #largemodel

Our team recently dove into Nvidia’s Omniverse and Isaac Sim to see how these platforms are shaping the future of robotics. Here’s what we found: Omniverse is carving out its niche in realistic scene and object modeling, rivaling giants like Unity and Unreal Engine. Its extensions for motion simulation, like PhysX and RTX Sensors, make it a powerful tool. However, the transition from simulation to real-world robotics isn't seamless. The communication methods between simulators and real sensors/motors often differ, posing a challenge. Gazebo, integrated with ROS, offers a smoother experience in this regard. Isaac Sim, on the other hand, is becoming a go-to for generating synthetic data, especially for LCBM (Large Content and Behavior Models) solutions. While it’s excellent for this purpose, it requires specific Nvidia-compatible hardware and lacks the abstraction level needed for broader compatibility. This makes it tough to transfer learnings across different hardware components. In our experiments, we’ve seen the effectiveness of synthetic data firsthand, training a traffic sign classifier with it and winning multiple awards in the AI GO competitions using Gym-Duckietown. Isaac Sim is promising, but understanding where to use it versus testing on hardware is key to leveraging its full potential. #Robotics #SyntheticData #Simulation #AutonomousSystems

The latest release of 𝐘𝐎𝐋𝐎𝐯𝟏𝟏 (Ultralytics) is here, and I couldn't wait to test it out using some intense FPV drone footage! 🎥✨ Here’s why YOLOv11 is set to revolutionize real-time object detection and AI-powered applications: Enhanced Feature Extraction: YOLOv11 offers a refined design that leads to better feature extraction, improving precision for detecting even the smallest objects. Faster & More Efficient: With a 22% reduction in parameters compared to YOLOv8m, YOLOv11 still boosts accuracy. This means quicker performance with no compromise on quality—ideal for fast-moving FPV footage. Perfect for Edge & Cloud: Whether you’re using it on edge devices, the cloud, or NVIDIA GPUs, YOLOv11 handles detection, segmentation, and pose estimation with ease, making it versatile for a range of applications. Real-Time Precision: Testing it on FPV drone footage was amazing! YOLOv11 ability to process data in real-time allows us to capture dynamic, fast-paced environments with higher accuracy than ever before. Certainly! Feel free to explore YOLOv11 and share your thoughts. As you dive into the latest advancements in AI and computer vision, don’t hesitate to ask for assistance or discuss any related topics. Happy testing, and let’s keep driving innovation forward together! #ai #computervision #yolo #ultralytics #deeplearning #dataScience #fpv #student #university #engineer