RoomStructureNetLineScorer is a Python implementation of my paper, "RoomStructNet: Learning to Rank Non-Cuboidal Room Layouts From Single View", published in 2021. This project focuses on a deep learning model that ranks non-cuboidal room layouts from a single-view image, enabling accurate estimation of complex indoor structures. It is designed for applications in computer vision, 3D scene understanding, and architectural analysis.

• Room Layout Ranking: Learns to rank non-cuboidal room layouts based on single-view images.
• Line-Based Scoring: Utilizes line segments to evaluate and score room structure hypotheses.
• Deep Learning Architecture: Employs a neural network to process visual and geometric cues.
• Visualization Support: Includes tools to visualize predicted layouts and scoring results.

1. Clone the repository:

   git clone https://github.com/xzhang311/RoomStructureNetLineScorer.git

2. Navigate to the project directory:

   cd RoomStructureNetLineScorer

3. Install dependencies:

   pip install -r requirements.txt

Run the main script to process an input image and rank room layouts:

python main.py --input path/to/image.jpg --output path/to/results

Use the --help flag for detailed configuration options:

python main.py --help

• Python 3.8+
• Libraries: PyTorch, NumPy, OpenCV, Matplotlib (listed in requirements.txt)

This project implements the methodology described in:

• "RoomStructNet: Learning to Rank Non-Cuboidal Room Layouts From Single View", CoRR, 2021. arXiv:2110.00644

If you find this work useful, please cite it using the following BibTeX:

@article{DBLP:journals/corr/abs-2110-00644,
  author       = {Xi Zhang and
                  Chun{-}Kai Wang and
                  Kenan Deng and
                  Tomas F. Yago Vicente and
                  Himanshu Arora},
  title        = {RoomStructNet: Learning to Rank Non-Cuboidal Room Layouts From Single
                  View},
  journal      = {CoRR},
  volume       = {abs/2110.00644},
  year         = {2021},
  url          = {https://arxiv.org/abs/2110.00644},
  eprinttype   = {arXiv},
  eprint       = {2110.00644},
  timestamp    = {Tue, 27 Feb 2024 16:41:39 +0100},
  biburl       = {https://dblp.org/rec/journals/corr/abs-2110-00644.bib},
  bibsource    = {dblp computer science bibliography, https://dblp.org}
}

Contributions are welcome! Please fork the repository, create a feature branch, and submit a pull request with your enhancements.

This project is licensed under the MIT License. See the LICENSE file for details.

This work is based on the research presented in the 2021 paper and leverages open-source computer vision and deep learning frameworks.