#196 · Primary category： Computer Vision
cosmo-edge
Production-grade C++ edge AI engine for video analytics and on-device VLM across Sophon, Rockchip RKNN, and x86, with visual orchestration, real-time OSD, events, and reproducible benchmarks.
Project last updated：08/28/26
GitHub Stars
621
Forks
152
Contributors
8
License
Apache-2.0
Why we included this project
Shipping video analytics to NPU-equipped edge boards usually means stitching together vendor SDKs, model conversion scripts, and custom glue code for each new device. CosmoEdge is a C++17 engine that collapses that whole path into one orchestration layer, covering Sophon BM1688/CV186X, Rockchip RK3576/RV1126B, and plain x86. Beyond inference serving, it handles model import, visual pipeline orchestration, real-time OSD overlays, alarm handling, and event delivery, so the result is a deployable application rather than a bare model endpoint. Reproducible multi-platform benchmark reports and a Docker-based x86 quick start let you evaluate the workflow on a laptop before committing to specific edge hardware. For teams building on-device vision products, the appeal is one codebase that targets several NPU vendors without locking you into a single runtime.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
opencv
Open Source Computer Vision Library
RuView
π RuView turns commodity WiFi signals into real-time spatial intelligence, vital sign monitoring, and presence detection — all without a single pixel of video.
PaddleOCR
Turn any PDF or image document into structured data for your AI. A powerful, lightweight OCR toolkit that bridges the gap between images/PDFs and LLMs. Supports 100+ languages.
MinerU
Transforms complex documents like PDFs and Office docs into LLM-ready markdown/JSON for your Agentic workflows.
tesseract
Tesseract Open Source OCR Engine (main repository)