zhuixunforever

why not multiply ray step size when volume rendering.

5

Hi, thanks for this great code. I have a puzzle, in your code, https://github.com/apchenstu/mvsnerf/blob/6e0111188ff92b13229482a95a2fb4eaddbc2971/renderer.py#L22 the 'sigma' not multiply step size 'dist', which is different in Ep(8) in your paper. By...

a small error in train.py

3

https://github.com/jnulzl/PRNet/blob/bbdc3c95266aa4ce6379cd343121616244964872/train.py#L39 this should be 'label', instead of 'img', like as: label_array = np.array(label,dtype=np.float32) by the way, thanks for your train code!

@yxlijun hai, sir, thank you your good job! how many iters did you use when you trained the voc dataset, and get [Pelee_VOC.pth](https://drive.google.com/open?id=16HparGAVhxTDByi5RylYCkxLZYducK9j)??? 1、160000? like ./configs/Pelee_VOC.py ? 2、and with 8...

A small gap between the pre-training model you provide and your paper in the loop of the RobotCar dataset

8

Thank you for your work! I download your pre-training model and eval it by evat.py This is LOOP2(2014-06-23-15-36-04) output: > Error in translation: median 4.7186 m, mean 8.5358 m Error...

Can you share me the depth map of DTU test sets

4

How do you get a high-resolution depth map? Mvsnet seems to provide only 128 * 160 depth maps. Can you share the high-resolution depth map with me? Thank you very...

In table 2 in your paper, you provide validation accuracy against ground truth depth maps with different models.  and I notice you divide depth_interval when calculating [l1loss](https://github.com/tejaskhot/unsup_mvs/blob/0d6c96a00d4247bbdceccfdbb002d7d649c6ae3d/code/unsup_mvsnet/model.py#L359) and [less_one_accuracy](https://github.com/tejaskhot/unsup_mvs/blob/0d6c96a00d4247bbdceccfdbb002d7d649c6ae3d/code/unsup_mvsnet/model.py#L371)....

The Cambridge label used for training is missing.

2

Hi, Thanks for the code. When I try to retrain your network, I noticed that you didn't seem to provide Cambridge label for training. The Cambridge label currently provided appears...

When I run this code, it showed that it was downloading the Face Alignment Network(FAN), so I downloaded it by the Web, but the tar I downloaded can't be open...

Hi, thanks for the wonderful work. Whether you use the foreground mask when calculating the F1-score of TaT, just like using the IDR mask when calculating the chamfer distance of...