ViVid-1-to-3: Novel View Synthesis with Video Diffusion Models