We propose a simple yet effective method to learn to segment new indoor
scenes from video frames: State-of-the-art methods trained on one dataset, even
as large as the SUNRGB-D dataset, can perform poorly when applied to images
that are not part of the dataset, because of the dataset bias, a common
phenomenon in computer vision. To make semantic segmentation more useful in
practice, one can exploit geometric constraints. Our main contribution is to
show that these constraints can be cast conveniently as semi-supervised terms,
which enforce the fact that the same class should be predicted for the
projections of the same 3D location in different images. This is interesting as
we can exploit general existing techniques developed for semi-supervised
learning to efficiently incorporate the constraints. We show that this approach
can efficiently and accurately learn to segment target sequences of ScanNet and
our own target sequences using only annotations from SUNRGB-D, and geometric
relations between the video frames of target sequences.
Beschreibung
[1904.12534] Casting Geometric Constraints in Semantic Segmentation as Semi-Supervised Learning
%0 Generic
%1 stekovic2019casting
%A Stekovic, Sinisa
%A Fraundorfer, Friedrich
%A Lepetit, Vincent
%D 2019
%K 2019 segmentation semi-supervised
%T Casting Geometric Constraints in Semantic Segmentation as
Semi-Supervised Learning
%U http://arxiv.org/abs/1904.12534
%X We propose a simple yet effective method to learn to segment new indoor
scenes from video frames: State-of-the-art methods trained on one dataset, even
as large as the SUNRGB-D dataset, can perform poorly when applied to images
that are not part of the dataset, because of the dataset bias, a common
phenomenon in computer vision. To make semantic segmentation more useful in
practice, one can exploit geometric constraints. Our main contribution is to
show that these constraints can be cast conveniently as semi-supervised terms,
which enforce the fact that the same class should be predicted for the
projections of the same 3D location in different images. This is interesting as
we can exploit general existing techniques developed for semi-supervised
learning to efficiently incorporate the constraints. We show that this approach
can efficiently and accurately learn to segment target sequences of ScanNet and
our own target sequences using only annotations from SUNRGB-D, and geometric
relations between the video frames of target sequences.
@misc{stekovic2019casting,
abstract = {We propose a simple yet effective method to learn to segment new indoor
scenes from video frames: State-of-the-art methods trained on one dataset, even
as large as the SUNRGB-D dataset, can perform poorly when applied to images
that are not part of the dataset, because of the dataset bias, a common
phenomenon in computer vision. To make semantic segmentation more useful in
practice, one can exploit geometric constraints. Our main contribution is to
show that these constraints can be cast conveniently as semi-supervised terms,
which enforce the fact that the same class should be predicted for the
projections of the same 3D location in different images. This is interesting as
we can exploit general existing techniques developed for semi-supervised
learning to efficiently incorporate the constraints. We show that this approach
can efficiently and accurately learn to segment target sequences of ScanNet and
our own target sequences using only annotations from SUNRGB-D, and geometric
relations between the video frames of target sequences.},
added-at = {2020-01-09T08:52:07.000+0100},
author = {Stekovic, Sinisa and Fraundorfer, Friedrich and Lepetit, Vincent},
biburl = {https://www.bibsonomy.org/bibtex/296750d8710ab52985cfeb6cc1a440faa/analyst},
description = {[1904.12534] Casting Geometric Constraints in Semantic Segmentation as Semi-Supervised Learning},
interhash = {56933142342f304cb57bc20198f9bcc5},
intrahash = {96750d8710ab52985cfeb6cc1a440faa},
keywords = {2019 segmentation semi-supervised},
note = {cite arxiv:1904.12534Comment: To be presented at WACV 2020},
timestamp = {2020-01-09T08:52:07.000+0100},
title = {Casting Geometric Constraints in Semantic Segmentation as
Semi-Supervised Learning},
url = {http://arxiv.org/abs/1904.12534},
year = 2019
}