cipher982/wine-images-126k

Wine Images Dataset 126K

A comprehensive dataset of 107,821 wine bottle images linked to the Wine Text Dataset 126K. This companion dataset provides high-quality wine bottle images for computer vision, multimodal machine learning, and wine recognition tasks.

Dataset Description

This dataset contains wine bottle images scraped from wine retailer websites. Each image is linked to detailed wine information (descriptions, pricing, categories, regions) via stable IDs that connect to the companion text dataset.

Key Features

• 107,821 wine bottle images in high resolution
• Stable linking to companion text dataset via image_id
• Clean naming: Images named as wine_XXXXXX.jpg matching text dataset IDs
• Quality images: Average 57KB per image, various resolutions
• Complete coverage: 98% of wines from text dataset have corresponding images

Dataset Structure

{
  "image_id": "wine_000001",           # Links to cipher982/wine-text-126k
  "image": <PIL.Image>,                # Wine bottle image
  "wine_name": "Dom Perignon Vintage 2008"  # Wine name for reference
}

Companion Dataset

This image dataset is designed to work with:

• cipher982/wine-text-126k: Text descriptions, prices, categories, regions

Usage

Basic Loading

from datasets import load_dataset

# Load the image dataset
image_dataset = load_dataset("cipher982/wine-images-126k")

# Load the companion text dataset
text_dataset = load_dataset("cipher982/wine-text-126k")

# Access images and text
images = image_dataset["train"]
texts = text_dataset["train"]

# Example: Get image and text for same wine
wine_id = "wine_000001"
wine_image = images.filter(lambda x: x["image_id"] == wine_id)[0]["image"]
wine_text = texts.filter(lambda x: x["id"] == wine_id)[0]

Multimodal Usage

import pandas as pd
from datasets import load_dataset

# Load both datasets
images = load_dataset("cipher982/wine-images-126k")["train"]
texts = load_dataset("cipher982/wine-text-126k")["train"]

# Convert to DataFrames for easy joining
df_images = images.to_pandas().set_index('image_id')
df_texts = texts.to_pandas().set_index('id')

# Join datasets on wine ID
df_multimodal = df_texts.join(df_images, how='inner')

print(f"Multimodal dataset: {len(df_multimodal):,} wines with both text and images")

# Example: Access wine with both image and description
wine = df_multimodal.iloc[0]
print(f"Name: {wine['name']}")
print(f"Description: {wine['description'][:100]}...")
print(f"Price: ${wine['price']}")
wine['image'].show()  # Display the wine bottle image

Computer Vision Tasks

from datasets import load_dataset
import torch
from torchvision import transforms

# Load dataset
dataset = load_dataset("cipher982/wine-images-126k")["train"]

# Preprocessing for computer vision models
transform = transforms.Compose([
    transforms.Resize((224, 224)),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.485, 0.456, 0.406],
                        std=[0.229, 0.224, 0.225])
])

# Process images
def preprocess(example):
    example["image"] = transform(example["image"])
    return example

dataset = dataset.map(preprocess)

Data Quality

• Image Count: 107,821 wine bottle images
• Coverage: 98.0% of wines from text dataset have images
• File Format: JPEG images
• Average Size: 57KB per image
• Total Size: ~5.8GB
• Naming: Consistent wine_XXXXXX.jpg format
• Quality: High-resolution product photos from wine retailers

Use Cases

Computer Vision

• Wine Classification: Classify wines by bottle shape, label, region
• Brand Recognition: Identify wine producers from bottle images
• Quality Assessment: Analyze bottle condition and presentation
• Object Detection: Detect wine bottles in complex scenes

Multimodal Learning

• Image-Text Matching: Match wine descriptions to bottle images
• Caption Generation: Generate wine descriptions from bottle images
• Visual Question Answering: Answer questions about wine bottles
• Recommendation Systems: Visual and textual wine recommendations

Research Applications

• Food & Beverage Analysis: Study wine packaging and branding trends
• Cultural Studies: Analyze wine bottle design across regions
• Marketing Research: Study visual elements in wine presentation
• Computer Vision Benchmarks: Large-scale wine image classification

Dataset Statistics

Image Coverage by Region

Region	Images	Coverage
other	84,127	79.5%
california	10,687	98.2%
france	4,753	98.2%
italy	4,254	98.4%

Image Coverage by Wine Category

Category	Images	Coverage
red_wine	60,893	97.9%
other	29,732	97.5%
white_wine	25,701	97.9%
rosé	2,658	98.0%
dessert	2,469	98.0%
sparkling	1,568	97.6%

Ethical Considerations

• Data Source: Images collected from public wine retailer websites
• Privacy: No personal information in images
• Commercial Use: Please respect original retailers' intellectual property
• Attribution: Images represent retailer product photography
• Quality: Images reflect commercial wine presentation standards

Technical Details

File Organization

wine-images-126k/
├── images/
│   ├── wine_000000.jpg
│   ├── wine_000001.jpg
│   └── ... (107,821 images)
├── image_metadata.json
└── README.md

Metadata Format

{
  "image_id": "wine_000001",
  "filename": "wine_000001.jpg",
  "original_filename": "lmgmud1xsenlouwpzysc.jpg",
  "file_size": 47234
}

Linking with Text Dataset

Images are linked to text data via stable image_id fields:

# Text dataset (cipher982/wine-text-126k)
{
  "id": "wine_000001",
  "name": "Dom Perignon Vintage 2008",
  "description": "Complex champagne with...",
  "image_id": "wine_000001"  # Links to this dataset
}

# Image dataset (cipher982/wine-images-126k)
{
  "image_id": "wine_000001",  # Same ID links back to text
  "image": <wine bottle image>,
  "wine_name": "Dom Perignon Vintage 2008"
}

Citation

If you use this dataset in your research, please cite:

@dataset{wine_images_126k,
  title={Wine Images Dataset 126K},
  author={David Rose},
  year={2025},
  url={https://huggingface.co/datasets/cipher982/wine-images-126k}
}

Also cite the companion text dataset:

@dataset{wine_text_126k,
  title={Wine Text Dataset 126K},
  author={David Rose},
  year={2025},
  url={https://huggingface.co/datasets/cipher982/wine-text-126k}
}

License

This dataset is released under the Creative Commons Attribution 4.0 International License (CC-BY-4.0).

You are free to:

• 🔄 Share — copy and redistribute the material in any medium or format
• 🔧 Adapt — remix, transform, and build upon the material for any purpose, even commercially

Under the following terms:

• 📝 Attribution — You must give appropriate credit and indicate if changes were made

Data Collection Notice: The underlying wine bottle images were collected from publicly available retailer websites for research purposes under fair use. This dataset compilation, stable ID system, and organized structure represent our original contribution covered by this license.

Users should respect the intellectual property rights of the original wine bottle photography and retailer content.