Image Segmentation with TensorFlow

This project demonstrates how to build an image segmentation pipeline using TensorFlow and TFRecords.
It covers dataset loading, preprocessing, batching, and visualization of input images, labels, and predictions.

📌 Features

• Load training and validation datasets from TFRecords.
• Parse and decode image/label data into TensorFlow tensors.
• Preprocess images (reshape, normalize, cast).
• Create optimized tf.data.Dataset pipelines for training and testing.
• Display input images, ground-truth labels, and predictions with Matplotlib.
• Ready-to-extend into a full segmentation model (U-Net, FCN, etc.).