Skip to content

Latest commit

 

History

History
49 lines (39 loc) · 2.94 KB

README.md

File metadata and controls

49 lines (39 loc) · 2.94 KB
Raw

MARL: [M]ulti-scale [A]rchetype [R]epresentation [L]earning and Clustering for Building Energy Estimation

Recent Updates:

  • 08/08/23: Our work is accepted to ICCV 2023 Workshop: 1st Computer Vision Aided Architectural Design (CVAAD) Workshop.

Citation:

If our work is useful or relevant to your research, please kindly recognize our contributions by citing our paper:

@InProceedings{Zhuang_2023_ICCV,
    author    = {Zhuang, Xinwei and Huang, Zixun and Zeng, Wentao and Caldas, Luisa},
    title     = {MARL: Multi-scale Archetype Representation Learning for Urban Building Energy Modeling},
    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops},
    month     = {October},
    year      = {2023},
    pages     = {1565-1572}
}

Overview:

This repository is the implementation code of the paper "MARL: [M]ulti-scale [A]rchetype [R]epresentation [L]earning and Clustering for Building Energy Estimation" (iccvw 2023, poster).

Frame_2

We present Multi-scale Archetype Representation Learning (MARL), a method designed to automate local building archetype construction through representation learning. Our proposed method addresses the aforementioned challenges by refining the essential elements of building archetypes for Urban Building Energy Modeling. This is a learning-based pipeline for representing and clustering buildings in our urban environment. Our research can be used in building energy estimation and can significantly save computing time.

Dataset:

  • For footprints and their meta info, please refer to this directory.
  • For building energy consumption data, all rights are reserved by the Lawrence Berkeley National Lab. Please contact authors for more detailed information.

Requirements:

Before running our data generation and annotation pipeline, you can activate a conda environment where Python Version >= 3.7:

conda create --name [YOUR ENVIR NAME] python = [PYTHON VERSION]
conda activate [YOUR ENVIR NAME]

then install all necessary packages:

pip install -r requirements.txt

Train:

To run training of our model, please refer to this notebook, or run the following command:

python train.py

Archetype Clustering:

To get latent representation and run clustering, please refer to this notebook.