Arxiv paper - Align3R: Aligned Monocular Depth Estimation for Dynamic Videos | AI Breakdown | Podwise