Master_Thesis/dataset.py

import os
import random
from PIL import Image, ImageFilter
from io import BytesIO
import torch
from torch.utils.data import Dataset
import torchvision.transforms as transforms


class ImageDataset(Dataset):
    def __init__(self, folder, label, transform, is_train=False, aug_prob=0, jpeg_qf=None, gaussian_blur=None):
        self.folder = folder
        self.label = label
        self.transform = transform
        self.is_train = is_train
        self.aug_prob = aug_prob
        self.jpeg_qf = jpeg_qf
        self.gaussian_blur = gaussian_blur

        self.images = [
            os.path.join(folder, img) for img in os.listdir(folder)
            if os.path.isfile(os.path.join(folder, img))
        ]
        self.jpeg_choices = [90, 75, 50]
        self.blur_choices = [1.0, 2.0, 3.0]

    def __len__(self):
        return len(self.images)

    def __getitem__(self, idx):
        image_path = self.images[idx]
        try:
            image = Image.open(image_path).convert("RGB")

            # === TRAIN MODE: apply JPEG/blur randomly ===
            if self.is_train:
                if random.random() < self.aug_prob:
                    qf = random.choice(self.jpeg_choices)
                    from io import BytesIO
                    buffer = BytesIO()
                    image.save(buffer, format="JPEG", quality=qf)
                    buffer.seek(0)
                    image = Image.open(buffer).convert("RGB")

                if random.random() < self.aug_prob:
                    sigma = random.choice(self.blur_choices)
                    image = image.filter(ImageFilter.GaussianBlur(radius=sigma))

            # === EVAL MODE: apply deterministic JPEG/blur if given ===
            else:
                if self.jpeg_qf:
                    from io import BytesIO
                    qf = random.choice(self.jpeg_qf)
                    buffer = BytesIO()
                    image.save(buffer, format="JPEG", quality=qf)
                    buffer.seek(0)
                    image = Image.open(buffer).convert("RGB")

                if self.gaussian_blur:
                    sigma = random.choice(self.gaussian_blur)
                    image = image.filter(ImageFilter.GaussianBlur(radius=sigma))

            image = self.transform(image)
            return image, self.label

        except Exception as e:
            print(f"Skipping corrupted image {image_path}: {e}")
            raise IndexError




# class ImageDataset(Dataset):
#     def __init__(self, folder, label, transform, jpeg_qf=None, gaussian_blur=None):
#         self.folder = folder
#         self.label = label
#         self.transform = transform
#         self.jpeg_qf = jpeg_qf
#         self.gaussian_blur = gaussian_blur
#         self.images = [
#             os.path.join(folder, img) for img in os.listdir(folder)
#             if os.path.isfile(os.path.join(folder, img))
#         ]

#     def __len__(self):
#         return len(self.images)

#     def __getitem__(self, idx):
#         image_path = self.images[idx]
#         try:
#             image = Image.open(image_path).convert("RGB")

#             # --- JPEG Compression ---
#             if self.jpeg_qf:
#                 buffer = BytesIO()
#                 qf = random.choice(self.jpeg_qf)
#                 image.save(buffer, format="JPEG", quality=qf)
#                 buffer.seek(0)
#                 image = Image.open(buffer).convert("RGB")

#             # --- Gaussian Blur ---
#             if self.gaussian_blur:
#                 sigma = random.choice(self.gaussian_blur)
#                 image = image.filter(ImageFilter.GaussianBlur(radius=sigma))
            
#             image = self.transform(image)
#             return image, self.label
#         except Exception as e:
#             print(f"Skipping corrupted image {image_path}: {e}")
#             raise IndexError