#157157th of 168 in ML frameworks and libraries
videopipe
C++ pipeline framework for video analysis with pluggable inference nodes
- Stars
- 3k
- License
- Apache-2.0
- Last commit
- Feb 2026
- Last release
- Apr 2024
- Language
- C++
2,731 on 21 Apr 20262,993 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
VideoPipe is a C++17 framework for building video analysis apps by linking independent nodes: stream sources, inference, tracking, behavior analysis, OSD, recording and outputs. Inference defaults to OpenCV::DNN, with TensorRT, PaddleInference, ONNXRuntime or multimodal LLM services (Ollama, vLLM, OpenAI-compatible) as alternatives. It ships over 40 sample pipelines and positions itself as a lighter alternative to DeepStream and mxVision.
Who it is for: C++ engineers building RTSP/video analytics pipelines for security or traffic
Strengths
- Runs CPU-only by default via OpenCV::DNN; CUDA and TensorRT are opt-in CMake flags
- Reads and pushes UDP, RTSP, RTMP and file streams, with Kafka or socket data output
- Includes tracking (IOU, SORT) and line-crossing and parking behavior analysis nodes
- Over 40 sample pipelines, plus a live pipeline status board
Weaknesses
- Must be built from source with CMake; no Docker image, last release 2024-04
- Model files and test data are downloaded from Google Drive or Baidu Drive
- Code for Cambricon, Rockchip and Ascend backends is not provided
- Self-rated medium performance; you write C++ to parse each model's output
What it needs
- GPU optional
- Needs OpenCV >= 4.6, GStreamer 1.14.5, GCC >= 7.5, CMake, CUDA (optional), TensorRT (optional), Paddle Inference (optional), ONNX Runtime (optional), Kafka (optional)
- Models: OpenCV::DNN, ONNX, TensorRT, PaddleInference, ONNXRuntime
Quick answers
Is videopipe open source?
Yes. Its code is on GitHub under the Apache-2.0 license.
Does videopipe need a GPU?
No. It runs on a CPU and can use a GPU when you have one.
Also in ML frameworks and libraries
See all 168| Rank | Project | Score |
|---|---|---|
| 1 | transformersPython library defining open models for text, vision, audio and multimodal tasks | 94 out of 100 |
| 2 | pytorchPython tensor library with GPU support and tape-based autograd | 93 out of 100 |
| 3 | ultralyticsPython library and CLI for training and running YOLO vision models | 93 out of 100 |
| #4 | pytorch-image-modelsPyTorch library of image models, pretrained weights and training scripts | 93 out of 100 |
| #5 | supervisionPython toolkit for loading, annotating and counting computer vision detections | 92 out of 100 |