| --- |
| task_categories: |
| - feature-extraction |
| - image-feature-extraction |
| - text-to-image |
| language: |
| - en |
| tags: |
| - ai |
| - architecture |
| - infrastructure |
| - genai |
| size_categories: |
| - 1K<n<10K |
| license: other |
| --- |
| |
| # Dataset Summary |
|
|
| This is a curated collection of floorplan images sourced from across the internet. It is intended for research in architectural AI, layout generation, and urban scene understanding. |
|
|
| - **Data format:** Image files with associated integer labels. |
| - **Sources:** Publicly available images from various web sources (This dataset is one unified collections). |
| - **Purpose:** Educational and research use. |
|
|
| ## Dataset Structure |
| The dataset follows the standard Hugging Face `Image` format: |
|
|
| | Column | Type | Description | |
| | :--- | :--- | :--- | |
| | `image` | `Image` | The visual data (decoded PIL image). | |
| | `label` | `int64` | Category ID (specific labels to be defined). | |
|
|
| ## Quick Start |
| To use this dataset in Python: |
|
|
| ```python |
| from datasets import load_dataset |
| |
| # Load the dataset |
| dataset = load_dataset("wheres-my-python/floorplans-cityscapes") |
| |
| # See a sample |
| print(dataset['train'][0]) |
| |
| |
| |
| |
| |
| |
