AyeshaIrshad1337/ImageCompression

This project involve training a sparse autoencoder using TensorFlow to compress and then reconstruct images

SparseAutoencoderImageCompression

This project demonstrates image compression and reconstruction using a sparse autoencoder built with TensorFlow.

File Structure

• data/: Contains data loading scripts and sample images.
• models/: Defines the sparse autoencoder model.
• utils/: Includes utility scripts for visualizing the results.
• main.py: Main script to load data, train the model, and visualize results.
• requirements.txt: Lists the required Python packages.

How to Run

1. Install the required packages:

   pip install -r requirements.txt
   

2. Place your sample images in the data/sample_images/ directory.

3. Run the main script to train the model and visualize results:

   python train.py
   

Description

The project trains a sparse autoencoder to compress and reconstruct images. The autoencoder consists of an encoder that compresses the image and a decoder that reconstructs it. The sparsity constraint encourages the model to use fewer neurons, effectively learning a more compact representation of the input data.