Skip to content

Latest commit

 

History

History
109 lines (84 loc) · 3.28 KB

INSTALL.md

File metadata and controls

109 lines (84 loc) · 3.28 KB
Raw

Set up the python environment

conda create -n e2ec python=3.7
conda activate e2ec

# install pytorch, the cuda version is 11.1
# You can also install other versions of cuda and pytorch, but please make sure # that the pytorch cuda is consistent with the system cuda

pip install torch==1.7.1+cu110 torchvision==0.8.2+cu110 torchaudio==0.7.2 -f https://download.pytorch.org/whl/torch_stable.html

pip install Cython==0.28.2
pip install -r requirements.txt

Compile cuda extensions

ROOT=/path/to/e2ec
cd $ROOT/network/backbone/DCNv2-master
# please check your cuda version and modify the cuda version in the command
export CUDA_HOME="/usr/local/cuda-11.1"
bash ./make.sh

Maybe you will encounter some build errors. You can choose a plan :

  1. You can look for another implementation of DCN-V2 and compiled successfully.
  2. You can set cfg.model.use_dcn as False. This may result in a slight drop in accuracy.
  3. You can install mmcv, and replace 352 line of network/backbone/dla.py as from mmcv.ops import ModulatedDeformConv2dPack as DCN, replace the deformable_groups in 353 line as deform_groups.

Set up datasets

Cityscapes

  1. Download the Cityscapes dataset (leftImg8bit_trainvaltest.zip) from the official website.

  2. Download the processed annotation file cityscapes_anno.tar.gz.

  3. Organize the dataset as the following structure:

    ├── /path/to/cityscapes
│   ├── annotations
│   ├── coco_ann
│   ├── leftImg8bit
│   ├── gtFine

  4. Create a soft link:

    ROOT=/path/to/e2ec
cd $ROOT/data
ln -s /path/to/cityscapes cityscapes

Kitti

  1. Download the Kitti dataset from the official website.

  2. Download the annotation file instances_train.json and instances_val.json from Kins.

  3. Organize the dataset as the following structure:

    ├── /path/to/kitti │ ├── testing │ │ ├── image_2 │ │ ├── instance_val.json │ ├── training │ │ ├── image_2 │ │ ├── instance_train.json

  4. Create a soft link:

     ROOT=/path/to/e2ec
 cd $ROOT/data
 ln -s /path/to/kitti kitti

Sbd

  1. Download the Sbd dataset at here.
  2. Create a soft link: 
    ROOT=/path/to/e2ec
cd $ROOT/data
ln -s /path/to/sbd sbd

COCO

  1. Download the Sbd dataset at here.

  2. Organize the dataset as the following structure:

    ├── /path/to/coco
│   ├── annotations
│   │   ├── instances_train2017.json
│   │   ├── instances_val2017.json
│   │   ├── image_info_test-dev2017.json
│   ├── train2017
│   ├── val2017
│   ├── test2017

  3. Create a soft link:

    ROOT=/path/to/e2ec
cd $ROOT/data
ln -s /path/to/coco coco