Installation

Docker

We recommend using docker images if docker or other container runtimes e.g. singularity is available on your devices.

We maintain a prebuilt image at dockerhub:

yangxue2docker/yx-tf-det:tensorflow1.13.1-cuda10-gpu-py3

Note

For 30xx series graphics cards (cuda11), please download image from tensorflow-release-notes according to your development environment, e.g. nvcr.io/nvidia/tensorflow:20.11-tf1-py3

Manual configuration

This repository is developed and tested with ubuntu 16.04, python 3.5 (anaconda recommend), tensorflow-gpu 1.13, cuda 10.0, opencv-python 4.1.1.26, tqdm 4.54.0, Shapely 1.7.1, tfplot 0.2.0 (optional). If docker is not available, we provide detailed steps to install the requirements by apt and pip.

Note

For 30xx series graphics cards (cuda11), we recommend this blog to install tf1.xx