README.md

metadata

license: mit

Clothes Brand Logos Dataset

Description:

The Clothes Brand Logos Dataset is a collection of images featuring logos from four popular clothing brands: Adidas, Allen Solly, Puma, and US Polo. The dataset includes both real and fake versions of these logos, providing a comprehensive resource for training and evaluating logo detection models.

Categories:

1. Adidas
2. Allen Solly
3. Puma
4. US Polo

Dataset Structure:

• Training Data:

  • The training data consists of eight subfolders directly representing the image categories.
  • Each subfolder contains real and fake logo images for a specific brand.

• Test Data:

  • The test data mirrors the structure of the training data, with eight subfolders representing each category.
  • Each subfolder contains real and fake logo images for a specific brand.

Folder Structure:

|-- Training
|   |-- Adidas_Real
|   |-- Adidas_Fake
|   |-- Allen_Solly_Real
|   |-- Allen_Solly_Fake
|   |-- Puma_Real
|   |-- Puma_Fake
|   |-- US_Polo_Real
|   |-- US_Polo_Fake
|
|-- Test
|   |-- Adidas_Real
|   |-- Adidas_Fake
|   |-- Allen_Solly_Real
|   |-- Allen_Solly_Fake
|   |-- Puma_Real
|   |-- Puma_Fake
|   |-- US_Polo_Real
|   |-- US_Polo_Fake

Usage:

• This dataset can be used for training and evaluating logo detection models, particularly those aimed at detecting fake logos.
• Researchers and practitioners can leverage this dataset to develop robust algorithms for identifying counterfeit logos and protecting brand integrity.