gdlg / panoramic-depth-estimation Goto Github PK

I got the depth data from the. npy file, but I found that the nearer the point in the panoramic image, the greater the depth value. I want to know how you define the depth. thank you.

Thanks for your work! I tried to use your wights for predict indoor scenes, but get very bad result. Did you try to work with indoor?

I think

    def generate_image_left(self, img, disp):
        if self.params.equirectangular_mode:
            return bilinear_sampler_equirectangular(img, -disp, self.params.fov)
        else:
            return bilinear_sampler_1d_h(img, -disp, self.params.fov)
    def generate_image_right(self, img, disp):
        if self.params.equirectangular_mode:
            return bilinear_sampler_equirectangular(img, disp)
        else:
            return bilinear_sampler_1d_h(img, disp)

should be

    def generate_image_left(self, img, disp):
        if self.params.equirectangular_mode:
            return bilinear_sampler_equirectangular(img, -disp, self.params.fov)
        else:
            return bilinear_sampler_1d_h(img, -disp)
    def generate_image_right(self, img, disp):
        if self.params.equirectangular_mode:
            return bilinear_sampler_equirectangular(img, disp, self.params.fov)
        else:
            return bilinear_sampler_1d_h(img, disp)

Thank you.

This is my command line：

python monodepth_simple.py --image_path ~/007998.png --output_path my_output --checkpoint_path models/panoramic_checkpoints/mixed_warp/model-180000 --input_height 256 --input_width 1024

The following error occurred：

Traceback (most recent call last):
File "monodepth_simple.py", line 16, in
import numpy as np
File "/home/nest/anaconda3/lib/python3.5/site-packages/numpy/init.py", line 142, in
from . import core
File "/home/nest/anaconda3/lib/python3.5/site-packages/numpy/core/init.py", line 57, in
from . import numerictypes as nt
File "/home/nest/anaconda3/lib/python3.5/site-packages/numpy/core/numerictypes.py", line 111, in
from ._type_aliases import (
File "/home/nest/anaconda3/lib/python3.5/site-packages/numpy/core/_type_aliases.py", line 63, in
_concrete_types = {v.type for k, v in _concrete_typeinfo.items()}
File "/home/nest/anaconda3/lib/python3.5/site-packages/numpy/core/_type_aliases.py", line 63, in
_concrete_types = {v.type for k, v in _concrete_typeinfo.items()}
AttributeError: 'tuple' object has no attribute 'type'

how to fix it?

Hi,

Thank you for this amazing work. I wanted to try your inference code on the proposed validation Carla dataset using one of your trained models. However when doing that, I get bad results that have nothing to do with the scores reported in your paper. For instance, when testing using the model in mixed_warp folder, I get the following results:

Abs. rel. & Sq. rel. & RMSE & RMSE log. & Depth acc < 1.25
0.810 & 67.533 & 8.520 & 1.427 & 0.051

The same thing goes for carla_warp and carla models
Is there something I am missing ? Note that I used the same evaluation code available with your dataset.
I will be grateful if you give some insights about what might cause the problem.

Best regards

The output of your program has only one image and one .npy file. How to get the predicted depth value of the image?

Hi.
I have questions about datasets used for training.
In your paper, the best performance was achieved by using CARLA & Mapillary datasets.
However, CARLA datasets which you uploaded, do not have stereo pair images.
How the training proceeds using those datasets??

Also, I wonder how the rectilinear images were transformed to equi rectangular images.
Any codes or reference links will be helpful.
Thank you.

like this:
Connecting to hades.ext.dur.ac.uk (hades.ext.dur.ac.uk)|129.234.103.162|:443... failed: Connection refused.