Community Projects

See what the community is building with reComputer.

inteintegrity2026-10-19

Build a Pose Estimation App with Codex on reComputer AI R2140

Build, deploy, and test a Hailo-8 pose estimation WebUI on reComputer AI R2140 with Codex, from SSH inspection to MP4 playback and download.

CodexCVPose Estimation
GitHub →
inteintegrity2026-10-16

Build an Edge AI App with Codex on reComputer RK3576

Learn to use Codex to inspect a reComputer RK3576 over SSH, build and deploy an NPU-powered video detection WebUI, and verify the complete application on real hardware using a repeatable, prompt-driven workflow.

CodexEdge AICV
GitHub →
inteintegrity2026-09-29

Qwen3-1.7B Long-Context Benchmark on RK1820

Test first-turn response, long-context retrieval, and multi-turn memory with Qwen3-1.7B running locally on reComputer RK3576 and an RK1820 accelerator.

Long ContextLLMBenchmark
GitHub →
inteintegrity2026-09-29

RK1820 Vision Story Studio

Upload an image and turn it into a postcard, short story, scene description, or creative prompts with a local Qwen2.5-VL-3B model on reComputer RK3576 and RK1820.

Image UnderstandingVLMCreative AI
GitHub →
inteintegrity2026-09-18

Multimodal Retail AI Demo Center on reComputer RK3588

A local retail AI showcase combining smart-shelf vision, inventory-aware voice interaction, four-stream RKNN detection, and sales conversation analysis on reComputer RK3588.

RetailMultimodal AICV
GitHub →
William Zhang2026-09-28

One-Hand Pinch 3×3 Puzzle Game Based on MediaPipe Hand Landmarker

Real-time detection of 21 single-hand landmarks and pinch state using the MediaPipe Tasks Hand Landmarker model mapping "thumb + index finger pinch" directly to "drag / release" semantics — the player draws the ROI and plays the puzzle in front of the camera using only pinch gestures, with no keyboard, mouse, or touch screen required in the entire interaction chain.

CVHand Tracking
Github →
inteintegrity2026-09-28

Qwen3-4B Long-Context Benchmark on RK1828

Explore first-turn response, needle-in-a-haystack retrieval, and multi-turn memory with Qwen3-4B running locally on reComputer RK3576 and an RK1828 accelerator.

Long ContextLLMBenchmark
GitHub →
William Zhang2026-09-27

9-Filter Dual-Hand Pinch Demo with MediaPipe Hand Landmarker

Real-time detection of 21 hand key points and pinch state for both hands on CPU (XNNPACK-accelerated) using the MediaPipe Tasks Hand Landmarker model. "Both hands pinching simultaneously" maps directly to a quadrilateral ROI, and 9 image filters are applied live to the pixels inside the ROI.

CVHand TrackingMediaPipe
Github →
Seeed Studio2026-09-24

Build an Interruptible Conversational Voice AI Terminal at the Edge

Build a voice terminal that listens and speaks locally, supports natural barge-in, and connects to either an OpenAI-compatible cloud endpoint or an on-device conversation model.

Voice AIEdge AILLMAEC
Seeed →
inteintegrity2026-09-22

Three accelerators on one RK3576: built-in NPU vs Hailo-8 vs RK182x

The same YOLO26n on the reComputer RK3576's built-in NPU, on a Hailo-8 module and on an RK182x module, compared on single-stream and 8-stream FPS: Hailo-8 is the fastest on a single stream, RK182x is the best at 8 streams.

CVBenchmarkEdge AI
GitHub →
Hanzo Huang2026-09-23

Build a Real-Time Queue Counter with YOLO11 on reComputer RK3576

Turn a video or RTSP stream into a live queue monitor that counts people only inside the area you define, powered by YOLO11 on the RK3576 NPU.

CVObject DetectionPeople Counting
GitHub →
Hanzo Huang2026-09-21

Zone Alert with YOLO11 on reComputer RK3576

Monitor a camera, video, or RTSP stream and record an event when a person or car enters a user-defined zone, with YOLO11 inference accelerated by the RK3576 NPU.

CVObject DetectionVideo Surveillance
GitHub →
William Zhang2026-09-22

Build a 3D Human Pose Estimation system using reComputer RK3576

On the RK3576, a system has been deployed for capturing human body joint points and football movement paths and performing 3D reconstruction during football dribbling. Through YOLOv8n, YOLOv8n-pose, and Mathplotlib demonstration graph drawing algorithms, it achieves the capture and display of human posture and football movement trajectories.

CVPose EstimationObject Detection
Github →
William Zhang2026-09-22

Build an NVR System Based on Frigate on the reComputer RK3576

Deploy Frigate NVR on the reComputer RK3576 with one-click Docker deployment, enabling hardware-accelerated decoding and RKNN inference.

NVRCVEdge AI
Github →
Parker Hu2026-09-20

RK3588 VPU and NPU Hardware Acceleration Benchmark

Reproducible MPP benchmarks for NV12-to-MJPG/H.264/H.265 hardware encoding, MJPG-to-YUV420 hardware decoding, and RKNN acceleration of YOLOv8n, YOLOv8s, and YOLOv8m inference on the reComputer RK3588.

VPUNPUBenchmark
Download →
Parker Hu2026-09-22

Understanding the Rockchip RKNN NPU Software Stack

Learn how RKNN-Toolkit2, RKNN Runtime, the RKNPU driver, and Rockchip NPU hardware work together, including model conversion, version compatibility, CNN, Transformer, and LLM deployment paths, and systematic troubleshooting on RK3576 and RK3588.

RKNNNPUEdge AI
GitHub →
William Zhang2026-09-20

Egg Counting and Crack Detection Based on reComputer RK3576

The egg counting and crack detection system based on reComputer RK3576 simultaneously performs egg counting and crack/dirt detection on a conveyor belt line. Through Conveyor orchestration and IoU association, it outputs the position, count, and damage condition of each egg.

YOLODamage DetectionEgg Counting
GitHub →
William Zhang2026-09-20

Intelligent Face Recognition System on the reComputer R2135

Uses YOLOv8n + SCRFD-10G + ArcFace MobileFaceNet on a reComputer (Hailo-8 NPU) to detect, align, and cluster pedestrians and faces in video, automatically building a face album organized by person that can be browsed from a minimal web page.

CVFace RecognitionYOLO
GitHub →
William Zhang2026-09-20

Real-time People-Attribute-Flow Analysis Web App on reComputer R + Hailo-8

Uses YOLOv8n pedestrian detection combined with Person Attribute ResNet attribute. recognition to perform offline video analysis on the reComputer R, with real-time。browser playback and statistics.

Person AttributePeople CountingYOLOEdge AICV
GitHub →
Hanzo Huang2026-09-14

PPE Area Detection with Raspberry Pi 5 and Hailo-8

Check hardhats, safety vests, and masks inside a user-defined area with Hailo-8 accelerated detection and a browser dashboard on Raspberry Pi 5.

CVObject DetectionPPE
GitHub →
inteintegrity2026-09-09

License Plate Recognition with Raspberry Pi 5 and Hailo-8

Detect license plates with Tiny-YOLOv4 and recognize numeric plate text with LPRNet using Hailo-8 accelerated inference on Raspberry Pi 5.

CVLicense Plate RecognitionObject Detection
GitHub →
inteintegrity2026-09-09

YOLO11 Quantization Comparison on reComputer RK3576

Convert YOLO11n from ONNX to RKNN and compare FP16, INT8, and W4A16 model size, inference speed, runtime memory, and COCO accuracy on reComputer RK3576.

CVBenchmarkQuantization
GitHub →
inteintegrity2026-09-03

Real-Time Pull-Up Counter on reComputer RK3576

A YOLO11 pose application that uses the RK3576 NPU to detect a person, draw the body skeleton and bounding box, and count pull-ups in real time from video, camera, or RTSP input.

CVPose EstimationFitness
Github →
Hanzo Huang2026-09-01

Building Real-Time Traffic Analytics on Raspberry Pi 5 with Hailo-8

Run local traffic analytics on Raspberry Pi 5 with Hailo-8. Count vehicles in selected areas, track their movement, and measure directional line crossings from one browser dashboard.

CVEdge AITraffic
Github →
Parker Hu2026-08-25

Deploying Viseron on the reComputer R2245

Benchmarks continuous Viseron recording on the reComputer R2245 and validates Hailo-8 integration.

NVRBenchmark
Other →
inteintegrity2026-08-18

Pothole Detection with Raspberry Pi 5 and Hailo-8

Detect road potholes from video using a YOLOv8n HEF model on Hailo-8, with ROI road-area filtering, temporal gap-fill tracking, and a browser MJPEG dashboard on Raspberry Pi 5.

CVObject DetectionRoad Safety
GitHub →
inteintegrity2026-08-11

Retail Shelf Detection with Raspberry Pi 5 and Hailo-8

Detect shelf products, track inventory changes, and monitor pickup and low-stock events from a browser dashboard with Hailo-8 accelerated inference on Raspberry Pi 5.

CVObject DetectionRetail
GitHub →
Parker Hu2026-07-29

RK3576 VPU and NPU Hardware Acceleration Benchmark

Benchmark results for MPP hardware-accelerated YUV420 to MJPG/H264/H265 encoding, MJPG to YUV420 decoding, and RKNN acceleration of YOLOv8n, YOLOv8s, and YOLOv8m inference on the reComputer RK3576.

VPUNPUBenchmark
Download →
inteintegrity2026-07-28

Airborne Object Detection on reComputer RK3576

VisDrone-trained YOLOv8s for aerial object detection on RK3576, supporting 11 airborne classes including pedestrian, vehicle, and cyclist.

CVObject DetectionDrone
Github →
Hanzo Huang2026-07-28

Building a Private Home Assistant Voice Assistant on the RK3576

Build a private, local Home Assistant voice assistant on the Rockchip RK3576. The stack combines Whisper for speech recognition, Piper for speech synthesis, openWakeWord for wake-word detection, and Qwen for local conversation—all accelerated by the RK3576 NPU.

LLMHome AssistantIoT
Github →
William Zhang2026-07-23

Multi-scenario Fall Detection Based on reComputer RK3576

Fall behavior detection in multi-scenario and multi-target situations using the YOLOv8n‑pose model, with inference acceleration powered by the reComputer RK3576, and simple information feedback via a web page.

CVPose EstimationFall Detection
Github →
Seeed Studio2026-07-22

Vehicle Obstacle Detection on reComputer RK3576

Real-time detection of vehicles, pedestrians, two-wheelers, and traffic signs in driving footage using YOLOv8n on the RK3576 NPU, with on-device inference accelerated by the 2 TOPS NPU; draw "obstacle risk zone" polygons on a web UI, trigger warnings for any target that enters a region, and emit annotated JPEG snapshots.

CVObject DetectionEdge AI
Github →
Parker Hu2026-07-22

YOLOv8 Real-time Inference CPP Project with reComputer RK3576

A multi-source input, multi-stream output, YOLOv8 real-time inference project based on RK3576, RKNN, MPP, RGA, and ZLMediaKit.

CVYOLOMPP
Github →
Parker Hu2026-07-17

Football Player Tracking Analysis Using reComputer RK3576

Leverage reComputer RK3576 RGA 2D hardware acceleration and VPU MPP acceleration. Run the YOLOv8n model to detect football players and achieve real-time analysis.

YOLOTrackingFootball
Github →
William Zhang2026-05-14

OpenClaw on reComputer - Privacy-First Emotion AI Agent

A privacy-first AI companion that detects your emotions via webcam and responds with the right tone—built on reComputer with OpenClaw.

Emotion DetectionOpenClawYOLO
hackster →
Seeed Studio2026-03-09

Object Detection with YOLO11 on reComputer-RK

Industrial CV solution for RK3588/RK3576: NPU-accelerated YOLO11 & MJPEG streaming.

YOLOCVObject Detection
Wiki →
seeed studio2025-12-16

Gun detection with Frigate NVR on R2000

Real-time AI firearm detection on reComputer R using YOLO11 and Hailo.

CVNVRYOLO
Wiki →
Seeed Studio2024-08-15

Deploy NVBlox with Orbbec Camera on Jetson AGX Orin

Isaac ROS NVBlox: GPU-powered real-time 3D mapping for robots.

Isaac ROSnvbloxRGB-D
Wiki →