/tensorrt-zoo

openpose, yolov3 with tiny-tensorrt

Primary LanguageCuda

This repo is no longer maintained

tensorrt-zoo

common computer vision models and some useful tools base on tiny-tensorrt.

Roadmap

  • openpose 🔥🔥🔥 --- 2019.10.17
  • yolov3

Quick Start

for run sample you need to install opencv and TensorRT

git clone --recursive https://github.com/zerollzeng/tensorrt-zoo.git
mkdir build && cd build && cmake -D PYTHON_API ON .. && make

for yolo3 sample see docs/yolov3.md for openpose sample see docs/openpose.md

System requirements

CUDA version >= 10.0 is fully test

TensorRT 6.0+, 5.x you need to reset old commit(not recommend)

OPENCV version >= 3.0, only use for read/write images and some basic image processing. so version 2.x might work but have not been tested.

About License

most of the openpose post processing was copy and modify slightly from official openpose repo,maybe you should distribute it under openpose's license, so as for some other code. for the other I wrote, You just DO WHAT YOU WANT TO.