An OBS plugin for removing background in portrait images (video), making it easy to replace the background when recording or streaming.

The Triton backend for the ONNX Runtime.

The Next-Gen Database for AI—an infrastructure designed for data and AI. As the MySQL of the AI era.

C++ implementations of YOLOv8/YOLO26, RT-DETR-V2(RTDETR), MobileSAM, and NanoSAM on TensorRT, ONNX Runtime, and RKNN.

ONNX Runtime for the Robot Operating System (ROS), works on ROS1 and ROS2

Classification wrappers for 4 Deep Learning Embedded Runtimes for Music Information Retrieval, plus utilities to convert a TensorFlow Model

一个 NTFS 文件恢复工具，支持 MFT 扫描、签名搜索、智能覆盖检测和高性能文件恢复

Real-time SAM2 segmentation on edge devices - 40x faster C++ inference with ONNX Runtime for iOS/Android deployment

lightweight inference engine/runtime

High-performance C++17 engine for real-time, stateful log anomaly detection. Uses a multi-tiered system combining heuristics, statistical Z-scores, and ONNX machine learning to find threats. Features flexible alerting (JSON, Syslog, HTTP) and live configuration reloading for operational maturity.

GPU-accelerated local semantic code index for Claude Code. CUDA-powered embeddings, HNSW vector search, tree-sitter AST chunking, MCP server — sub-3ms queries, fully on your machine.

Trying to build ONXX runtime execution engine

High-perf C++ AI/ML inference engine with ONNX Runtime & LibTorch. CPU default, GPU opt-in. CLI + REST API, cross-platform via CMake/Docker. Auto-fetch models/assets, CUDA runtime check, CI merge gating, future multi-model & GPU training support.