README.md

---
license: mit
task_categories:
- image-classification
- zero-shot-image-classification
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
dataset_info:
  features:
  - name: image
    dtype: image
  - name: label
    dtype:
      class_label:
        names:
          '0': rocher_chocolate
          '1': milka_chocolate
          '2': kinder_chocolate
          '3': toblerone_white
          '4': toblerone_black
          '5': lays_classic
          '6': lays_chill
          '7': pringles_original
          '8': pringles_paprika
          '9': nestle_water
          '10': sanpellegrino_water
          '11': redbull_energydrink
          '12': monster_energydrink
          '13': heinz_ketchup
          '14': heinz_mayo
          '15': barilla_pesto
          '16': barilla_pomodoro
          '17': barilla_lasagne
          '18': loreal_shampoo
          '19': dove_soap
          '20': sensodyne_toothpaste
          '21': colgate_toothpaste
          '22': sensodyne_mouthwash
          '23': nivea_rollon
          '24': rexona_spray
          '25': dove_spray
          '26': nivea_baby
          '27': johnson_baby
  - name: image_hash
    dtype: string
  - name: product_name
    dtype: string
  - name: sample_id
    dtype: string
  splits:
  - name: train
    num_bytes: 437784366.375
    num_examples: 10357
  - name: test
    num_bytes: 193054870.0
    num_examples: 4920
  download_size: 630652408
  dataset_size: 630839236.375
---

# Micro Market Experience (MIMEX) Product Recognition Dataset

## Dataset Description

The Micro Market Experience (MIMEX) Dataset is a comprehensive collection of product images tailored for retail and consumer product classification tasks. This dataset encompasses 15,277 images across 28 product categories, including chocolates, snacks, beverages, personal care items, and food products.

## Dataset Features

- **Total Images**: 15,277
- **Number of Classes**: 28
- **Splits**: Train (10,357) / Test (4,920)
- **Image Format**: PNG, RGB
- **Resolution**: Variable (original product images)

## Product Categories

The dataset includes the following 28 product categories:

- **0**: Rocher Chocolate
- **1**: Milka Chocolate
- **2**: Kinder Chocolate
- **3**: Toblerone White
- **4**: Toblerone Black
- **5**: Lays Classic
- **6**: Lays Chill
- **7**: Pringles Original
- **8**: Pringles Paprika
- **9**: Nestle Water
- **10**: Sanpellegrino Water
- **11**: Redbull Energydrink
- **12**: Monster Energydrink
- **13**: Heinz Ketchup
- **14**: Heinz Mayo
- **15**: Barilla Pesto
- **16**: Barilla Pomodoro
- **17**: Barilla Lasagne
- **18**: Loreal Shampoo
- **19**: Dove Soap
- **20**: Sensodyne Toothpaste
- **21**: Colgate Toothpaste
- **22**: Sensodyne Mouthwash
- **23**: Nivea Rollon
- **24**: Rexona Spray
- **25**: Dove Spray
- **26**: Nivea Baby
- **27**: Johnson Baby

## Dataset Structure

```
mimex-dataset/
├── train/           # Training images (10,357 samples)
├── test/            # Test images (4,920 samples)
└── metadata/        # Dataset information and mappings
```

## Data Fields

- **image**: PIL Image of the product
- **label**: Integer class label (0-27)  
- **image_hash**: Unique 12-character hash identifier
- **product_name**: Human-readable product name
- **sample_id**: Original sample identifier from source dataset

## How to Use

### Loading the Dataset

```
from datasets import load_dataset

# Load the full dataset
dataset = load_dataset("Anilot/MIMEX")

# Load specific split
train_dataset = load_dataset("Anilot/MIMEX", split="train")
test_dataset = load_dataset("Anilot/MIMEX", split="test")

print(f"Train samples: {len(train_dataset)}")
print(f"Test samples: {len(test_dataset)}")
```

### Basic Usage Example

```
import matplotlib.pyplot as plt
from datasets import load_dataset

# Load dataset
dataset = load_dataset("Anilot/MIMEX", split="train")

# Show first few samples
fig, axes = plt.subplots(2, 3, figsize=(12, 8))
for i, ax in enumerate(axes.flat):
    sample = dataset[i]
    ax.imshow(sample['image'])
    ax.set_title(f"{sample['product_name']}\nLabel: {sample['label']}")
    ax.axis('off')
plt.tight_layout()
plt.show()
```

## License

This dataset is released under the MIT License. Please cite appropriately if used in research.

## Citation

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

```
Tur, Anil Osman, et al. "Exploring Fine-grained Retail Product Discrimination with Zero-shot Object Classification Using Vision-Language Models." 2024 IEEE 8th Forum on Research and Technologies for Society and Industry Innovation (RTSI). IEEE, 2024.
```

DOI: [10.1109/RTSI61910.2024.10761839](https://doi.org/10.1109/RTSI61910.2024.10761839)

### BibTeX
```
@inproceedings{tur2024exploring,
  title={Exploring Fine-grained Retail Product Discrimination with Zero-shot Object Classification Using Vision-Language Models},
  author={Tur, Anil Osman and Conti, Alessandro and Beyan, Cigdem and Boscaini, Davide and Larcher, Roberto and Messelodi, Stefano and Poiesi, Fabio and Ricci, Elisa},
  booktitle={2024 IEEE 8th Forum on Research and Technologies for Society and Industry Innovation (RTSI)},
  pages={97--102},
  year={2024},
  organization={IEEE}
}
```