Cover Image

Saved in:
Bibliographic Details
Main Authors: Schulz, Daniel, Jonas, Tobias, Bayerl, Sebastian P.
Format: Recurso digital
Language:
Published: Zenodo 2025
Subjects:
Symbol Detection
Deep Learning
Building Plan Analysis
Document Analysis
Machine Learning
Artificial Intelligence
Construction Industry
Object Detection
Instance Segmentation
Blueprint Analysis
Bill of Materials
Architectural Plans
Online Access:https://doi.org/10.5281/zenodo.16749425
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • <p>The growing demand for housing has increased the need for intelligent systems in building planning to improve underlying processes and reduce the workload of professionals. A central aspect of creating such systems is the detection of symbols in building plans, which is essential for creating a bill of materials (BOM). This necessitates symbol data that reflects the wide variety of symbols used in practice. However, public datasets are not available, which led to an increased focus on methods for generating synthetic data. To address these issues, this dataset contains 2,000 distinct ELectrical INstallation symbols from Architectural plans (ELINA). Additionally, sample training, validation and test dataset splits are provided for model development.</p>

Similar Items