Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

I do not get the expected results #16

Open
mikeletxeberria opened this issue Jan 26, 2021 · 1 comment
Open

I do not get the expected results #16

mikeletxeberria opened this issue Jan 26, 2021 · 1 comment

Comments

@mikeletxeberria
Copy link

Hi, when I try to execute the code as it is, I can not achieve your results. The system and configuration I am using are the following ones:

OS: Windows 10
Conda env:

backcall=0.2.0=pypi_0
certifi=2020.6.20=py36_0
cudatoolkit=10.0.130=h5327add_6
cudnn=7.6.5=cuda10.0_0
cupy=6.0.0=py36h230ac6f_0
cupy-cuda102=9.0.0a2=pypi_0
cycler=0.10.0=pypi_0
easydict=1.9=pypi_0
fastrlock=0.5=pypi_0
intel-openmp=2020.3=h57928b3_311
joblib=0.17.0=pypi_0
kiwisolver=1.3.1=pypi_0
libblas=3.8.0=21_mkl
libcblas=3.8.0=21_mkl
liblapack=3.8.0=21_mkl
matplotlib=3.0.3=pypi_0
mkl=2020.4=hb70f87d_311
numpy=1.16.2=pypi_0
opencv-python=4.4.0.46=pypi_0
pillow=8.0.1=pypi_0
pip=20.2.4=py36haa95532_0
pyparsing=2.4.7=pypi_0
python=3.6.12=h5500b2f_2
python-dateutil=2.8.1=pypi_0
python_abi=3.6=1_cp36m
pyyaml=5.3.1=pypi_0
scikit-learn=0.23.2=pypi_0
scipy=1.5.4=pypi_0
setuptools=50.3.1=py36haa95532_1
six=1.15.0=pyh9f0ad1d_0
sklearn=0.0=pypi_0
sqlite=3.33.0=h2a8f88b_0
threadpoolctl=2.1.0=pypi_0
torch=1.4.0=pypi_0
torchvision=0.5.0=pypi_0
tqdm=4.54.0=pypi_0
vc=14.1=h0510ff6_4
vs2015_runtime=14.16.27012=hf0eaf9b_3
wheel=0.35.1=pyhd3eb1b0_0
wincertstore=0.2=py36h7fe50ca_0
zlib=1.2.11=h62dcd97_4

Command: python run.py -d options\kitti\kitti_default_configuration.yml

I have just changed the references to the KITTI dataset folder and the align_corners=True. The result is the following (even the ground truth is not printed as shown in the paper) :

map

Thank your for your help,

Mikel Etxeberria

@Huangying-Zhan
Copy link
Owner

Hi Mikel, could you check if the flow model is named correctly?
The flow model should be named as "network-default.pytorch" in the current implementation because there is a hard-coded setup which recognises the name.

# FIXME: hardcode for network-default.pytorch

# for free to join this conversation on GitHub. Already have an account? # to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants