Wonders of the World Image Classification
Image dataset containing 3846 photos of the new wonders of the world, each file representing a famous monument. Used to train multi-class image classification models.
3846 images divided into folders by marvel, JPEG/PNG formats (mainly JPEG)
CC0: Public Domain
Description
This dataset includes 3846 high-quality images, organized into folders corresponding to new wonders of the world, such as the Taj Mahal, the Great Wall of China, or the Colosseum in Rome. These images were collected from Google Images to allow visual recognition and classification of monuments.
What is this dataset for?
- Training computer vision models for multi-class image classification
- Development of AI tourism applications to identify sites and monuments
- Deep learning research on the recognition of images of cultural places
Can it be enriched or improved?
Yes, it is possible to add additional images to improve visual diversity or to create additional annotations (e.g. geolocation, season, angles of view). The dataset can also be remixed with other databases to increase the robustness of the models.
🔎 In summary
🧠 Recommended for
- Computer vision students
- AI developers
- Tourism projects
🔧 Compatible tools
- TensorFlow
- PyTorch
- FastAI
- OpenCV
💡 Tip
Pre-process images (resize, normalization) to optimize training.
Frequently Asked Questions
Can this dataset be used for the real-time recognition of a monument?
Yes, with a well-trained model, it is possible to use it for real-time recognition on mobile or web.
Does the dataset include images with different light conditions and viewing angles?
Mostly varied images collected on the web, which offers a natural diversity of angles and conditions.
Can I add my own images to enrich this dataset?
Yes, it is recommended that you annotate your images according to existing categories to maintain consistency and improve performance.
