In this challenge, you will tackle a **Supervised Image Denoising and Classification** task using a custom flower image dataset.
The dataset contains five flower classes:

1. **Daisy**
2. **Dandelion**
3. **Roses**
4. **Sunflowers**
5. **Tulips**

A portion of the dataset has been corrupted with **noise**, making classification harder. Your goal is to design a pipeline that:

1. **Removes noise** from the given images.
2. **Classifies** them into the correct flower category.

The **training set** contains both noisy and clean versions of the same images, enabling supervised learning for the denoising stage. The **test set** contains only noisy images, randomly ordered, with no clean references.

Your final submission will be a CSV file with two columns:

```
Images,Predicted_Classes
Image_Test_1.png,3
Image_Test_2.png,5
...
```

Where class labels are mapped as:

```
1 = Daisy
2 = Dandelion
3 = Roses
4 = Sunflowers
5 = Tulips
```

This competition is part of the **180DC Machine Learning SIG recruitment process**, aimed at testing your ability to work with noisy, real-world-like data while demonstrating strong problem-solving skills in computer vision.