Explore cutting-edge deep image prior techniques for solving inverse problems in computational imaging using MATLAB. This project implements untrained convolutional neural networks as natural image priors for tasks like denoising and super-resolution. The tutorial covers medical imaging (MRI/CT), satellite enhancement, and microscopy applications with complete MATLAB workflow: from handling corrupted inputs to implementing the optimization loop. Compare results with wavelet-based and total variation methods, highlighting advantages in texture preservation. Includes practical guidance on hyperparameter tuning and GPU acceleration for faster convergence.