Effective And Efficient Use Of Diffusion Models For Editing In Computer Vision