search

ewrfcas / MVSFormer

Codes of MVSFormer: Multi-View Stereo by Learning Robust Image Features and Temperature-based Depth (TMLR2023)
Apache License 2.0
181 stars 10 forks source link

DTU_aug_masks #12

Open Leroy0709 opened 1 year ago

Leroy0709 commented 1 year ago

请问这个 File "/root/MVSFormer/datasets/dtu_datasetms.py", line 310, in getitem img, depthhr, intrinsics_, depth_maskhr, offset_y, offset_x = self.final_crop(img, depth_hr, intrinsics, depth_mask_hr, File "/root/MVSFormer/datasets/dtu_dataset_ms.py", line 226, in final_crop offset_y = random.randint(0, h - crop_h) File "/root/miniconda3/lib/python3.8/random.py", line 248, in randint return self.randrange(a, b+1) File "/root/miniconda3/lib/python3.8/random.py", line 226, in randrange raise ValueError("empty range for randrange() (%d, %d, %d)" % (istart, istop, width)) ValueError: empty range for randrange() (0, -558, -558) 这里报错该怎么更改设置呢？

ewrfcas commented 1 year ago

你好，mask是没有用到的。

2023年3月26日 15:44，Leroy0709 @.***> 写道：

哪里能得到wmlce/DTU_MVSNet/DTU_aug_masks/irregular_mask*？

— Reply to this email directly, view it on GitHub https://github.com/ewrfcas/MVSFormer/issues/12, or unsubscribe https://github.com/notifications/unsubscribe-auth/AC34AE3BFFQNRWMSAXYUATTW57XUDANCNFSM6AAAAAAWIAUYHE. You are receiving this because you are subscribed to this thread.

ewrfcas commented 1 year ago

请问这个 File "/root/MVSFormer/datasets/dtu_datasetms.py", line 310, in getitem img, depthhr, intrinsics_, depth_maskhr, offset_y, offset_x = self.final_crop(img, depth_hr, intrinsics, depth_mask_hr, File "/root/MVSFormer/datasets/dtu_dataset_ms.py", line 226, in final_crop offset_y = random.randint(0, h - crop_h) File "/root/miniconda3/lib/python3.8/random.py", line 248, in randint return self.randrange(a, b+1) File "/root/miniconda3/lib/python3.8/random.py", line 226, in randrange raise ValueError("empty range for randrange() (%d, %d, %d)" % (istart, istop, width)) ValueError: empty range for randrange() (0, -558, -558) 这里报错该怎么更改设置呢？

这个似乎是使用了其他数据集训练 ，或者错改了image size setting?

McCartney2 commented 1 year ago

Traceback (most recent call last): File "test.py", line 580, in save_depth(testlist, config) File "test.py", line 203, in save_depth checkpoint = torch.load(config.resume) File "/root/miniconda3/lib/python3.8/site-packages/torch/serialization.py", line 608, in load return _legacy_load(opened_file, map_location, pickle_module, pickle_load_args) File "/root/miniconda3/lib/python3.8/site-packages/torch/serialization.py", line 777, in _legacy_load magic_number = pickle_module.load(f, pickle_load_args) EOFError: Ran out of input 这一步是测试集dtu出问题了吗？

ewrfcas commented 1 year ago

Traceback (most recent call last): File "test.py", line 580, in save_depth(testlist, config) File "test.py", line 203, in save_depth checkpoint = torch.load(config.resume) File "/root/miniconda3/lib/python3.8/site-packages/torch/serialization.py", line 608, in load return _legacy_load(opened_file, map_location, pickle_module, pickle_load_args) File "/root/miniconda3/lib/python3.8/site-packages/torch/serialization.py", line 777, in _legacy_load magic_number = pickle_module.load(f, pickle_load_args) EOFError: Ran out of input 这一步是测试集dtu出问题了吗？

似乎是加载权重出问题了

McCartney2 commented 1 year ago

processing /root/autodl-tmp/OUTPUT2/Temple, ref-view96, photo/geo/final-mask:0.08137015998363495/4.78707136153389e-07/0.0 Traceback (most recent call last): File "test.py", line 585, in pcd_filter(testlist) File "test.py", line 560, in pcd_filter pcd_filter_worker(scan) File "test.py", line 555, in pcd_filter_worker dynamic_filter_depth(pair_folder, scan_folder, out_folder, os.path.join(args.outdir, save_name)) File "test.py", line 497, in dynamic_filter_depth reproj_depth[~vis_mask.squeeze(2)] = 1 # [n v h w ] IndexError: too many indices for tensor of dimension 4 请问这个是因为显存不够吗？

maybeLx commented 1 year ago

processing /root/autodl-tmp/OUTPUT2/Temple, ref-view96, photo/geo/final-mask:0.08137015998363495/4.78707136153389e-07/0.0 Traceback (most recent call last): File "test.py", line 585, in pcd_filter(testlist) File "test.py", line 560, in pcd_filter pcd_filter_worker(scan) File "test.py", line 555, in pcd_filter_worker dynamic_filter_depth(pair_folder, scan_folder, out_folder, os.path.join(args.outdir, save_name)) File "test.py", line 497, in dynamic_filter_depth reproj_depth[~vis_mask.squeeze(2)] = 1 # [n v h w ] IndexError: too many indices for tensor of dimension 4 请问这个是因为显存不够吗？

你这个应该是代码跑的有问题，这里跑的是reprojection， 一个是confidence 大于某一个阈值，另一个则是 冲投影误差小于某个阈值，但是这里的的mask很低，应该是哪儿跑的有问题。

omnipotenttom commented 3 weeks ago

请问这个 File "/root/MVSFormer/datasets/dtu_datasetms.py", line 310, in getitem img, depthhr, intrinsics_, depth_maskhr, offset_y, offset_x = self.final_crop(img, depth_hr, intrinsics, depth_mask_hr, File "/root/MVSFormer/datasets/dtu_dataset_ms.py", line 226, in final_crop offset_y = random.randint(0, h - crop_h) File "/root/miniconda3/lib/python3.8/random.py", line 248, in randint return self.randrange(a, b+1) File "/root/miniconda3/lib/python3.8/random.py", line 226, in randrange raise ValueError("empty range for randrange() (%d, %d, %d)" % (istart, istop, width)) ValueError: empty range for randrange() (0, -558, -558) 这里报错该怎么更改设置呢？

这个似乎是使用了其他数据集训练 ，或者错改了image size setting? 作者您好，非常感谢您的工作！我遇到了一样的问题，我在使用dtu数据集时将您提供的Depths_raw解压到了原始的dtu数据集中然后开始运行代码image，结果 image