Ayke

Environment: Ubuntu 16.04, CUDA 10

The trickiest part is, OpenCV 4 only supports CUDA 10, not CUDA 10.1. So if you have installed CUDA 10.1, you have to get 10.0 instead.

Installation

Install required packages

sudo apt-get install gcc gfortran python-dev libopenblas-dev liblapack-dev cython

sudo apt-get install build-essential

Ayke

Installation

Environment: Ubuntu 16.04

Package requirement for ffmpeg

sudo apt-get update -qq && sudo apt-get -y install \
  autoconf \
  automake \
  build-essential \
  cmake \

Ayke

Driver install

To stop the currently running nvidia driver:

sudo systemctl isolate multi-user.target
sudo modprobe -r nvidia-drm

CUDA Installation

Ayke

OpenCV version: 4.3.0 Python version: 3.6

Prerequisite

Required libraries

sudo apt-get update
sudo apt-get install curl libssl-dev
sudo apt-get install ffmpeg

Ayke

First I assumed that you have already downloaded YOLO to /home/nvidia/darknet where nvidia is the host name.

YOLO installation

mavx options unavailable

CC: error: unrecognized command line option ‘-mavx’
CC: error: unrecognized command line option ‘-mavx2’
CC: error: unrecognized command line option ‘-msse3’
CC: error: unrecognized command line option ‘-msse4.1’
CC: error: unrecognized command line option ‘-msse4.2’

Ayke

Install Detectron 2

As it says here, https://github.com/facebookresearch/detectron2/blob/master/INSTALL.md we need to install a few modules before installing detectron.

The only part can be difficult is to install pytorch on TX2 since TX2 is aarch64 architecture, the installation instructions on pytorch website do not work. We need to install via this page: https://devtalk.nvidia.com/default/topic/1049071/jetson-nano/pytorch-for-jetson-nano/

To upgrade pyyaml, just pip3 install pyyaml --ignore-installed

Also you might want to keep using python3 and pip3 to avoid using python2

Ayke

Download source file from python website Goes with README

lsb_release problem:

subprocess.CalledProcessError: Command '('lsb_release', '-a')' returned non-zero exit status 1. Solution: sudo rm /usr/bin/lsb_release

Ayke

Prerequisites, i.e. the most important things

0. Time of writing: Jan 18, 2023, updated on Sep 22, 2024. The following assumes that you're trying to install CUDA on WSL2 Ubuntu.

1. Check support matrix first before you install any version of CUDA, because chances are the latest CUDA does not have cuDNN support yet, then you'll have to re-install older version if you found out later.

   https://docs.nvidia.com/deeplearning/cudnn/support-matrix/index.html#cudnn-cuda-hardware-versions

   At the time of writing, the latest cuDNN version is 8.7 and it supports CUDA 11.8.

2. Windows 10 must be build 20145 or later, or you should be on Windows 11.