Hello, I am very honored to see this great work. Using monocular depth to assist reconstruction is intuitively correct but actually difficult to do. Your idea is very interesting. I have a question, does this work support COLMAP's SfM, because MASt3R has high memory requirements and is difficult to run on my device. Have you tried to use the sparse depth map generated by COLMAP to help alignment? Can you give me some advice, thank you!