Add files using upload-large-folder tool

README.md

@@ -0,0 +1,144 @@
+# Colored MNIST Dataset
+
+A comprehensive dataset of MNIST digits with RGB colored backgrounds, designed for multi-objective classification tasks including digit recognition and background color classification.
+
+![Blue Example](images/blue_312.png) ![Green Example](images/green_8218.png) ![Red Example](images/red_8428.png)
+
+## Overview
+
+This dataset extends the classic MNIST digit dataset by adding colored backgrounds (blue, green, red) to create a multi-label classification challenge. Each image contains:
+- **Digit label**: 0-9 (digit recognition)  
+- **Color label**: blue, green, red (background color classification)
+- **Image format**: 28×28 RGB PNG files
+
+## Dataset Statistics
+
+### Training Set (8,835 samples)
+| Digit | Count | Digit | Count |
+|-------|-------|-------|-------|
+| 0     | 868   | 5     | 764   |
+| 1     | 1,000 | 6     | 893   |
+| 2     | 858   | 7     | 948   |
+| 3     | 909   | 8     | 835   |
+| 4     | 883   | 9     | 877   |
+
+### Test Set (10,000 samples)
+| Digit | Count | Digit | Count |
+|-------|-------|-------|-------|
+| 0     | 980   | 5     | 892   |
+| 1     | 1,135 | 6     | 958   |
+| 2     | 1,032 | 7     | 1,028 |
+| 3     | 1,010 | 8     | 974   |
+| 4     | 982   | 9     | 1,009 |
+
+### Color Distribution
+- **Blue**: ~2,946 training samples
+- **Green**: ~2,968 training samples  
+- **Red**: ~2,921 training samples
+
+## Directory Structure
+
+```
+colored_mnist/
+├── README.md
+├── colored_mnist.py          # HuggingFace dataset loader
+├── images/                   # Example images
+│   ├── blue_312.png         # Blue background, digit 3
+│   ├── green_8218.png       # Green background, digit 8  
+│   └── red_8428.png         # Red background, digit 4
+├── training/                 # Training data
+│   ├── 0/                   # Digit 0 samples
+│   │   ├── blue_1015.png
+│   │   ├── green_1000.png
+│   │   └── red_1028.png
+│   └── ...                  # Digits 1-9
+├── testing/                  # Test data
+│   ├── 0/                   # Digit 0 samples
+│   │   ├── blue_10.png
+│   │   ├── green_1084.png
+│   │   └── red_1001.png
+│   └── ...                  # Digits 1-9
+├── training.h5              # HDF5 format (optional)
+└── testing.h5               # HDF5 format (optional)
+```
+
+## Usage
+
+### Loading with HuggingFace Datasets
+
+```python
+from datasets import load_dataset
+
+# Load the dataset using the custom script
+dataset = load_dataset("FrankCCCCC/colored_mnist", trust_remote_code=True)
+
+print(f"Train samples: {len(dataset['train'])}")
+print(f"Test samples: {len(dataset['test'])}")
+
+# Access a sample
+sample = dataset['train'][0]
+print(f"Image shape: {sample['image'].shape}")  # PIL.Image: Raw Image with shape (28, 28, 3)
+print(f"Digit Label: {sample['digit_label']}")  # Integer: Digit class 0-9
+print(f"Color Label: {sample['color_label']}")  # String: Color class: 'blue', 'green', 'red'
+print(f"filename: {sample['filename']}")        # String: file name
+print(f"image_path: {sample['image_path']}")    # String: file path
+```
+
+## Applications
+
+This dataset is ideal for:
+
+1. **Multi-task Learning**: Train models to simultaneously classify digits and colors
+2. **Domain Adaptation**: Study how background colors affect digit recognition
+3. **Bias Analysis**: Investigate color-digit correlations and model robustness
+4. **Transfer Learning**: Pre-train on one task (digit/color) and fine-tune on another
+5. **Curriculum Learning**: Start with single-task then progress to multi-task
+6. **Attention Mechanisms**: Visualize what models focus on (digit vs. background)
+
+## Features
+
+- **Multi-label Classification**: Both digit and color labels for each image
+- **Balanced Colors**: Approximately equal distribution across blue, green, red
+- **Standard Format**: Compatible with PyTorch, TensorFlow, and HuggingFace
+- **PIL Integration**: Images loaded as PIL objects for easy preprocessing
+- **Metadata**: Includes original filename and image path information
+
+## Dataset Format
+
+Each sample contains:
+```python
+{
+    'image': PIL.Image,           # 28×28 RGB image
+    'digit_label': int,           # Digit class (0-9)
+    'color_label': str,           # Color class ('blue'|'green'|'red')
+    'filename': str,              # Original filename
+    'image_path': str             # Full path to image file
+}
+```
+
+## Requirements
+
+```bash
+pip install datasets pillow torch torchvision matplotlib numpy
+```
+
+## Citation
+
+```bibtex
+@misc{colorized-mnist,
+  title={Colorized MNIST},
+  author={Original MNIST Dataset},
+  year={2024},
+  url={https://github.com/jayaneetha/colorized-MNIST}
+}
+```
+
+## License
+
+This dataset follows the same license as the original MNIST dataset. Please cite both the original MNIST work and this colored extension when using this dataset.
+
+## Acknowledgments
+
+- Based on the original MNIST dataset by Yann LeCun et al.
+- Color augmentation creates additional research opportunities for multi-task learning
+- Compatible with modern deep learning frameworks and HuggingFace ecosystem

colored_mnist.py

@@ -0,0 +1,99 @@
+import os
+import datasets
+from datasets import Features, ClassLabel, Image, Value
+from PIL import Image as PILImage
+
+
+_DESCRIPTION = """\
+Colored MNIST is a dataset of MNIST digits with RGB colored backgrounds. 
+This dataset can be used for multi-objective classification tasks including:
+- Digit recognition (0-9)
+- Background color classification (blue, green, red)
+"""
+
+_CITATION = """\
+@misc{colorized-mnist,
+  title={Colorized MNIST},
+  url={https://github.com/jayaneetha/colorized-MNIST},
+}
+"""
+
+_HOMEPAGE = "https://github.com/jayaneetha/colorized-MNIST"
+
+_LICENSE = ""
+
+class ColoredMnist(datasets.GeneratorBasedBuilder):
+    """Colored MNIST dataset with multi-label classification."""
+
+    VERSION = datasets.Version("1.0.0")
+
+    def _info(self):
+        features = Features({
+            'image': Image(),
+            'digit_label': ClassLabel(num_classes=10, names=[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]),
+            'color_label': ClassLabel(num_classes=3, names=['blue', 'green', 'red']),
+            'image_path': Value('string'),
+            'filename': Value('string')
+        })
+
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=features,
+            homepage=_HOMEPAGE,
+            license=_LICENSE,
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        """Returns SplitGenerators."""
+        # The dataset path should be provided as a data_dir, or use current directory
+        data_dir = self.config.data_dir if self.config.data_dir else os.path.dirname(os.path.abspath(__file__))
+        print(f"data_dir: {data_dir}")
+        
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "data_dir": os.path.join(data_dir, "training"),
+                    "split": "train",
+                },
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.TEST,
+                gen_kwargs={
+                    "data_dir": os.path.join(data_dir, "testing"),
+                    "split": "test",
+                },
+            ),
+        ]
+
+    def _generate_examples(self, data_dir, split):
+        """Yields examples."""
+        idx = 0
+        raise NotImplementedError
+        
+        for digit in range(10):
+            digit_dir = os.path.join(data_dir, str(digit))
+            if not os.path.exists(digit_dir):
+                continue
+            
+            for filename in sorted(os.listdir(digit_dir)):
+                if filename.endswith('.png'):
+                    # Extract color from filename (e.g., "blue_1015.png" -> "blue")
+                    color = filename.split('_')[0]
+                    if color not in ['blue', 'green', 'red']:
+                        continue
+                    
+                    image_path = os.path.join(digit_dir, filename)
+                    
+                    # Load image as PIL Image
+                    pil_image = PILImage.open(image_path)
+                    
+                    yield idx, {
+                        'image': pil_image,
+                        'digit_label': digit,
+                        'color_label': color,
+                        'image_path': image_path,
+                        'filename': filename,
+                    }
+                    idx += 1