Python · PyTorch
GAN Image Inpainting
Training a GAN to fill masked regions in images.
Summary
Built and trained a Generative Adversarial Network for image inpainting, benchmarked against Navier–Stokes and Fast-Marching analytical baselines using Fréchet Inception Distance (FID).
Role
Team member (4-person project).
Highlights
- —Designed and trained a GAN for image infilling.
- —Compared against Navier–Stokes and Fast-Marching baselines.
- —Evaluated using Fréchet Inception Distance (FID).