Synthetic Satellite Data Generation

I'm looking for a professional who can generate synthetic satellite data using stable diffusion and a GANs. The goal is to create a satisfactory dataset from around 10 Google images depicting water bodies (like rivers, beaches, lakes) with litter. This data will be used to train a pre-existing model. The generated dataset should ultimately improve the accuracy and efficiency of a yet-to-be-determined pre-trained algorithm on this new dataset.