Skip to content

Latest commit

 

History

History
133 lines (108 loc) · 5.38 KB

README.md

File metadata and controls

133 lines (108 loc) · 5.38 KB
Raw

[CVPR 2024] Semantic Line Combination Detector

Jinwon Ko, Dongkwon Jin and Chang-Su Kim.

overview

Official code for "Semantic Line Combination Detector" in CVPR 2024. [arxiv].

Dataset

Download the following datasets to root/Datasets/. SEL and SEL_Hard datasets are provided in here. NKL dataset is provided in here. A new dataset called CDL is available at here.

Installation

  1. Create conda environment:
$ conda create -n SLCD python=3.6 anaconda
$ conda activate SLCD
$ conda install pytorch==1.10.1 torchvision==0.11.2 torchaudio==0.10.1 cudatoolkit=11.3 -c pytorch -c conda-forge
$ pip install opencv-python==4.7.0.72
  1. If you want to get the performance of the paper, download our pre-trained model to root/Modeling/pretrained/ and preprocessed data for SEL, SEL_Hard, NKL(SL5K), and CDL datasets to root/Preprocessing/.

Directory structure

.                           # ROOT
├── Modeling                # directory for modeling
│   ├── Detector
|   |   ├── code            
│   ├── SLCD           
|   |   ├── code            
│   ├── pretrained          # pretrained model parameters      
|   |   ├── Detector      
|   |   |   ├── checkpoint_paper_SEL.pth
|   |   |   ├── checkpoint_paper_NKL.pth
|   |   |   ├── checkpoint_paper_CDL.pth
|   |   ├── SLCD   
|   |   |   ├── checkpoint_paper_SEL.pth
|   |   |   ├── checkpoint_paper_NKL.pth
|   |   |   ├── checkpoint_paper_CDL.pth    
├── Preprocessing           # directory for preprocessed data
│   ├── SEL                 
|   |   ├── pickle             
│   ├── SEL_Hard            
|   |   ├── pickle             
│   ├── NKL                 
|   |   ├── pickle             
│   ├── CDL                 
|   |   ├── pickle             
├── Datasets                # Dataset directory
│   ├── SEL                 # SEL dataset
|   |   ├── ICCV2017_JTLEE_gt_pri_lines_for_test
|   |   ├── ICCV2017_JTLEE_gtlines_all
|   |   ├── ICCV2017_JTLEE_images
|   |   ├── Readme.txt
|   |   ├── test_idx_1716.txt
|   |   ├── train_idx_1716.txt
│   ├── SEL_Hard            # SEL_Hard dataset
|   |   ├── data
|   |   ├── edge
|   |   ├── gtimgs
|   |   ├── images
|   |   ├── README
│   ├── NKL                 # NKL dataset
|   |   ├── Data
|   |   ├── train.txt
|   |   ├── val.txt  
│   ├── CDL                 # CDL dataset
|   |   ├── train           
|   |   |   ├── Images
|   |   |   ├── Labels
|   |   ├── test            
|   |   |   ├── Images
|   |   |   ├── Labels

Evaluation

Run with

cd root/Modeling/SLCD/code/
python main.py

Train

For training line detector

  1. Edit root/Modeling/Detector/code/config.py. Please modify run_mode to 'train'. Also, set the dataset you want to train (dataset_name).
  2. Run with
$ cd root/Modeling/Detector/code/
$ python main.py

For training SLCD

  1. Edit root/Modeling/SLCD/code/config.py. Please modify run_mode to 'train'. Also, set the dataset you want to train (dataset_name).
  2. Run with
$ cd root/Modeling/SLCD/code/
$ python main.py

Test

  1. If you want to evaluate a model you trained, edit root/Modeling/SLCD/code/config.py. Please modify run_mode to 'test'. Also, set the dataset you want to test (dataset_name).
  2. Run with
$ cd root/Modeling/SLCD/code/
$ python main.py

Results

  1. Semantic line detection

Semantic line detection

  1. Road lane detection

Road lane detection

  1. Composition-based image retrieval

Composition-based image retrieval

  1. Symmetric axis detection

Symmetric axis detection

  1. Vanishing point detection

Vanishing point detection

Semantic feature grouping results Semantic feature grouping