Skip to content
New issue

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

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

Already on GitHub? Sign in to your account

MaskFusion always behaving as if "-static" flag was set #8

Open
Eman7C7 opened this issue Jan 30, 2019 · 1 comment
Open

MaskFusion always behaving as if "-static" flag was set #8

Eman7C7 opened this issue Jan 30, 2019 · 1 comment

Comments

@Eman7C7
Copy link

Eman7C7 commented Jan 30, 2019

Hi,

I built MaskFusion following the build.sh script (but with CUDA 9.0).
I am trying to run it on the fr3_walking_halfsphere sequence. The only class I enabled for the segmentation is "person", which, if I understand correctly, should be filtered out for the registration. However, the software ignores completely the semantic segmentation. To confirm this I tried to run it with the "-static" flag and it behaves exactly the same. I also tried to precompute the masks (using offline_runner.py and passing them with "-maskdir"). The masks are computed correctly, but again MaskFusion completely ignores them. The terminal does not show anything unusual, except for the following messages:

Your GPU "GeForce GTX 1080" isn't in the ICP Step performance database, please add it
Your GPU "GeForce GTX 1080" isn't in the RGB Step performance database, please add it
Your GPU "GeForce GTX 1080" isn't in the RGB Res performance database, please add it
Your GPU "GeForce GTX 1080" isn't in the SO3 Step performance database, please add it

What am I missing?

@ryohachiuma
Copy link

ryohachiuma commented Mar 4, 2019

Did you specify SPECIAL_ASSIGNMENTS in MaskRCNN.py?
I think you should change line 72, 73 the following.

FILTER_CLASSES = ['person']
SPECIAL_ASSIGNMENTS = {'person': 255}

For the message,

 the ICP Step performance database, please add it

You should read ElasticFusion github README.

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

No branches or pull requests

2 participants