Image Denoising & Reconstruction Using Autoencoders (CIFAR-10 Deep Learning Project)

I developed a deep learning project focused on image reconstruction and denoising using Autoencoder neural networks trained on the CIFAR-10 dataset. The project includes three different architectures: Vanilla Autoencoder (Fully Connected Neural Network) Denoising Autoencoder (with Gaussian noise injection) Convolutional Autoencoder (CNN-based model for better image feature extraction) Key features of the project: Data preprocessing and normalization of CIFAR-10 dataset Implementation of multiple neural network architectures using PyTorch Image compression into a latent representation (feature embedding) Reconstruction of images from compressed representations Denoising capability using noisy input training Visualization of original vs reconstructed images This project demonstrates strong knowledge of deep learning, computer vision, and neural network design.