UniqAI/gaussian-splatting
1
0
Fork 0
mirror of https://github.com/graphdeco-inria/gaussian-splatting synced 2025-04-02 20:30:40 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

calculate loss one by one image for saving gpu memory

Browse Source
This commit is contained in:
liuzhi 2024-06-04 21:08:06 +08:00
parent 9e82ac7170
commit da44a55c40
1 changed files with 8 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
train.py
View File

@ -111,7 +111,14 @@ def training(dataset, opt, pipe, testing_iterations, saving_iterations, checkpoi
render_pkg = render(viewpoint_cam, gaussians, pipe, bg, return_depth=True, return_normal=True)
image, viewspace_point_tensor, visibility_filter, radii = render_pkg["render"], render_pkg["viewspace_points"], render_pkg["visibility_filter"], render_pkg["radii"]
# Loss
# Loss，计算渲染图像与真实图像之间的损失
# 替换为一张一张读取
# gt_image_path = viewpoint_cam.image_path
# gt_image = Image.open(gt_image_path)
# gt_image = (torch.from_numpy(np.array(gt_image)) / 255.0).permute(2, 0, 1)
# gt_image = gt_image.clamp(0.0, 1.0).to(viewpoint_cam.data_device)
gt_image = viewpoint_cam.original_image.cuda()
Ll1 = l1_loss(image, gt_image)
loss = (1.0 - opt.lambda_dssim) * Ll1 + opt.lambda_dssim * (1.0 - ssim(image, gt_image))