Modified Swiss Dwellings Dataset

Dataset Composition and Features:

• Comprehensive Building Coverage: The dataset includes floor plans from both single-unit and multi-unit buildings, offering a broad spectrum of architectural layouts. This diversity is invaluable for training machine learning models that need to generalize across various building types.
• High-Resolution Floor Plans: Each floor plan in the MSD dataset is highly detailed, capturing the intricacies of Swiss architectural designs. The data includes precise geometries such as room layouts, wall structures, and other critical building elements.
• Derived from Swiss Dwellings Database: The MSD dataset is entirely derived from the Swiss Dwellings database. Ensuring that the data is grounded in real-world architectural standards and practices. This makes it an ideal resource for researchers and developers working on projects that require authentic building data.

Naming Convention and Dataset Split:

• Consistent Naming: The naming convention for the dataset follows the original Swiss Dwellings database. Ensuring consistency and ease of reference. The IDs used in the folders are directly inherited from the original dataset. Making it straightforward to cross-reference and integrate with other datasets.
• Train and Test Split: The dataset is split into distinct train and test sets based on the buildings from which the floor plans originate. This careful splitting ensures that there is no overlap between building identities in the train and test sets. As a result, all floor plans from a specific building are either entirely in the train set or the test set. Preventing data leakage and ensuring robust model evaluation.
• Enhanced Dataframe: To further aid in the development of machine learning algorithms. The dataset includes a cleaned, filtered, and modified Pandas dataframe. This dataframe contains all geometries derived from the original dataset.  Including details such as rooms, walls, and other structural elements. The dataframe is designed to be fully compatible with the unique floor plan IDs in both the train and test sets. Providing a solid foundation for advanced data analysis and algorithm development.

Potential Applications:

The Modified Swiss Dwellings dataset is an invaluable resource for a wide range of applications, including:

• Automated Floor Plan Generation: Train models to generate realistic floor plans based on existing building layouts.
• Architectural Analysis: Conduct detailed analysis of building structures, room configurations.  And spatial relationships within complex architectural designs.
• Image and Structure Extraction: Develop algorithms to extract and analyze structural features from floor plans.  Enhancing applications in computer vision and building information modeling (BIM).
• Graph-Based Analysis: Utilize the dataset for graph extraction and analysis. Enabling the study of spatial relationships and connectivity within building complexes.

Conclusion:

The Modified Swiss Dwellings (MSD) dataset represents a significant advancement in the availability of high-quality. ML-ready architectural data. Its extensive coverage, high resolution, and meticulous organization make it an ideal resource for researchers, developers, and architects looking to push the boundaries of floor plan generation. Building analysis, and related applications. With its robust train-test split and enhanced dataframe. The MSD dataset is poised to drive innovation in architectural machine learning and beyond.