rdk-yolo-toolkit端到端 YOLO 训练 → 地瓜 RDK X5 BPU 量化部署:Monkey Patch ONNX 导出 → hb_mapper 量化 → 板端 hbm_runtime 推理 → TROS HobotDnn ROS2 实时检测。支持 YOLOv5u~YOLOv13 和 YOLO26 全系列。内含工艺原理速查(why this ONNX, why this yaml)。
Install via ClawdBot CLI:
clawdbot install maxma615/rdk-yolo-toolkitGrade Fair — based on market validation, documentation quality, package completeness, maintenance status, and authenticity signals.
Accesses system directories or attempts privilege escalation
sudo chmodCalls external URL not in known-safe list
https://repo.anaconda.com/pkgs/mainAudited Oct 2, 2026 · audit v1.0
Generated Oct 2, 2026
A robotics team uses the RDK YOLO Toolkit to fine-tune a YOLOv8 segmentation model on custom factory-floor imagery, export it to ONNX, quantize with hb_mapper, and deploy to an RDK X5 board on a mobile robot. The toolkit's monkey-patch scripts and hb_mapper configs eliminate common operator and calibration pitfalls, yielding a real-time 30 FPS defect detector without cloud dependency.
An agritech startup leverages the toolkit to train YOLOv11 pose models for crop stalk counting and fruit ripeness detection on field-captured drone footage. After quantizing to the RDK X5 BPU, they run inference onboard solar-powered edge nodes, enabling real-time, low-power crop analytics even in remote areas with no connectivity.
A retail solutions provider deploys YOLOv8 OBB models to detect and classify products on store shelves using RDK X5-based smart cameras. The RDK YOLO Toolkit automates the training-to-deployment pipeline, including ROS2 TROS integration, so store associates receive live restocking alerts directly through a local dashboard without transmitting sensitive video to the cloud.
A manufacturing engineer uses the toolkit to train a YOLOv9 segmentation model to spot micro-cracks on PCB assemblies. After quantizing with hb_mapper and verifying cosine similarity >0.8, the model runs on an RDK X5 embedded in the conveyor belt, performing 100% inline inspection at high speed while rejecting defective units automatically.
A municipal technology office uses the RDK YOLO Toolkit to deploy YOLO26 detection and OBB models on RDK X5 edge boxes attached to traffic lights. The system tracks vehicle density and classifies traffic violations in real time. Native ROS2 HobotDnn integration simplifies streaming detection results to a central traffic management system.
Offer bespoke YOLO model training, quantization, and RDK X5 deployment as a service for clients needing edge AI but lacking in-house expertise. This includes dataset curation, model selection, optimization, and ROS2 integration, delivered as a fixed-scope project.
Sell pre-configured RDK X5 development kits preloaded with the RDK YOLO Toolkit, sample models, and step-by-step guides. Target makers, startups, and educators who want a turnkey edge AI platform, bundling hardware with a premium version of the toolkit that includes priority support and additional model templates.
Build a cloud-based pipeline that automates YOLO model training, quantization, and OTA deployment to fleets of RDK X5 edge devices. Customers upload their datasets, select model architectures, and the platform handles the entire workflow, including remote monitoring and performance analytics.
💬 Integration Tip
Always verify board-side environment (OS ≥3.5.0, dnn_node ≥2.6.1, hbm_runtime) before deployment using the provided check script. Ensure the HobotDnn class_num matches the cls_names_list and avoid using 'yolov8' parser for quantized models—use 'ultralytics_yolo' instead.
Scored Oct 2, 2026
Control remote Windows machines via SSH. Use when executing commands on Windows, checking GPU status (nvidia-smi), running scripts, or managing remote Windows systems. Triggers on "run on Windows", "execute on remote", "check GPU", "nvidia-smi", "远程执行", "Windows 命令".
Perform reverse lookup of gTLD domains hosted on a specified nameserver with optional filters by TLD and domain prefix length.
Essential curl commands for HTTP requests, API testing, and file transfers.
Connect to remote desktops via RDP, VNC, and SSH X11 with secure tunneling and troubleshooting.
Monitor web pages for content changes and get alerts. Track URLs, detect updates, view diffs. Use when asked to watch a website, track changes on a page, mon...
Comprehensive DevOps skill for CI/CD, infrastructure automation, containerization, and cloud platforms (AWS, GCP, Azure). Includes pipeline setup, infrastruc...