search

brownvc / diffdiffdepth

Official repository for CVPR 2021 paper "Differentiable Diffusion for Dense Depth Estimation from Multi-view Images"
http://visual.cs.brown.edu/diffdiffdepth/
GNU General Public License v3.0
63 stars 6 forks source link

Encountered an error #1

Open JimmyJune opened 3 years ago

JimmyJune commented 3 years ago

run cmd: python run_mvs.py --input_dir=/home/jimmy/oss/Project/video_sample/keyframes/VID_20210901_112934 --src_img=00001.jpeg --output_dir=./results

error: Traceback (most recent call last): File "run_mvs.py", line 179, in optimize(args.input_dir, args.src_img, args.output_dir) File "run_mvs.py", line 102, in optimize optim_param(optimizer_s) File "run_mvs.py", line 75, in optim_param torch.tensor(config.tile_max_pts), config.factor, config.loss_VGG, config.loss_grad, config.loss_smoothness) File "/home/jimmy/oss/Project/diffdiffdepth/loss.py", line 97, in loss_mvs gy) # gradient weights, y File "/home/jimmy/oss/Project/diffdiffdepth/LAHBPCG.py", line 486, in set_constraints self.AD = sx + sx_r + w + sy + sy_r RuntimeError: The size of tensor a (3840) must match the size of tensor b (3780) at non-singleton dimension 2

noobtoob4lyfe commented 1 year ago

I'm getting the same error. Any luck getting past this?

sjtu-medialab commented 1 year ago

I'm getting the same error. Any luck getting past this?