Hi author, nice work! When I use your code, I notice that your reference image gt_img and gt_img_torch come from the rendering of gt_scene. I wonder if it is possible to only use 0_gt.png as a reference image to guide the lay-out process. Because in practice, I tend to only have one correct image (without camera pose) instead of gt_scene. Feel free to correct me if I've misinterpreted anything.🤝
Hi author, nice work! When I use your code, I notice that your reference image
gt_img
andgt_img_torch
come from the rendering ofgt_scene
. I wonder if it is possible to only use0_gt.png
as a reference image to guide the lay-out process. Because in practice, I tend to only have one correct image (without camera pose) instead of gt_scene. Feel free to correct me if I've misinterpreted anything.🤝