search

JiayuYANG / CVP-MVSNet

Cost Volume Pyramid Based Depth Inference for Multi-View Stereo (CVPR 2020 Oral)
239 stars 33 forks source link

ERROR #30

Open liushifu12138 opened 3 years ago

liushifu12138 commented 3 years ago

Traceback (most recent call last): File "train.py", line 196, in train() File "train.py", line 137, in train loss = train_sample(sample, detailed_summary=do_summary) File "train.py", line 183, in train_sample loss.append(model_loss(depth_est_list[i], depth_gt.float(), mask)) File "/mnt/CVP-MVSNet-master/CVP_MVSNet/models/net.py", line 195, in sL1_loss return F.smooth_l1_loss(depth_est[mask], depth_gt[mask], reduction='mean') IndexError: The shape of the mask [1, 128, 160] at index 1 does not match the shape of the indexed tensor [1, 512, 640] at index 1

JiayuYANG commented 3 years ago

Hi liushifu12138,

Are you using the training dataset provided by MVSNet? The training set provided by MVSNet is different from the one I provide, since MVSNet and CVP-MVSNet have different structure. MVSNet take 512x640 image in and estimate depth map of 128x160 size. CVP-MVSNet take 128x160 image in and estimate depth map of the same resolution.

Cheers, Jiayu.

liushifu12138 commented 3 years ago

How to test on tank data set？ Can you share your code？please

wln19 commented 2 years ago

hello , can you test on tank data