tensorkelechi/arxiv-cstext · Datasets at Hugging Face

text stringlengths 0 89.3k
12 The comparison results are reported in Tab 5 SWNN
improves the accuracy of 3D shape recovery in most de
formation settings The accuracy improvement mainly de
pends on the segmentation result and the smoothing prop
erty of the sequence for more analysis see Supp Sec 3
Performance on Noisy and Missing Data We follow the
approaches in 30 to add noise and occlusion to the data
For missing data we first complete the measurement ma
trix by 9 and then run our method Fig 5 shows that our
method performs better than others in most cases3
5 Conclusion
In this paper we have proposed a spatialtemporal modeling
framework for NRSfM We proposed a Temporallysmooth
Procrustean Alignment module to tackle the rotation ambi
guity Furthermore we introduced a spatialweighted nu
clear norm model with the shape proxy strategy to ensure
the effectiveness of lowrank constraint in coping with se
vere spatiallyvariant deformations Our method achieves
superior 3D reconstruction performance across a wide range
of deformation sequences Moreover there is still much re
search to do in the future on more accurate trajectory seg
mentation and explicit modeling of complex deformation
Acknowledgements This research was supported in part
by the National Natural Science Foundation of China
62271410 and the Fundamental Research Funds for the
Central Universities
3See Supp for more tests on TPA SWNN and Missing DataReferences
1 Antonio Agudo Unsupervised 3d reconstruction and group
ing of rigid and nonrigid categories IEEE Trans Pattern
Anal Mach Intell 441519532 2020 1 2
2 Ijaz Akhter Yaser Sheikh Sohaib Khan and Takeo Kanade
Nonrigid structure from motion in trajectory space In Adv
Neural Inform Process Syst pages 4148 2008 1 2 6 7
3 Ijaz Akhter Yaser Sheikh and Sohaib Khan In defense of
orthonormality constraints for nonrigid structure from mo
tion In IEEE Conf Comput Vis Pattern Recog pages
15341541 2009 2
4 Benedicte Bascle and Andrew Blake Separability of pose
and expression in facial tracking and animation In Int Conf
Comput Vis pages 323328 1998 2
5 V olker Blanz and Thomas Vetter A morphable model for
the synthesis of 3d faces In Proceedings of the 26th an
nual conference on Computer graphics and interactive tech
niques pages 187194 1999 2
6 Stephen Boyd Neal Parikh Eric Chu Borja Peleato
Jonathan Eckstein et al Distributed optimization and sta
tistical learning via the alternating direction method of mul
tipliers Foundations and Trends in Machine learning 3
11122 2011 6
7 Matthew Brand A direct method for 3d factorization of non
rigid motion observed in 2d In IEEE Conf Comput Vis
Pattern Recog pages 122128 2005 2
8 Christoph Bregler Aaron Hertzmann and Henning Bier
mann Recovering nonrigid 3d shape from image streams
InIEEE Conf Comput Vis Pattern Recog pages 690696
2000 1 2 3
9 Ricardo Cabral Fernando De la Torre Jo ao P Costeira and
Alexandre Bernardino Unifying nuclear norm and bilinear
factorization approaches for lowrank matrix decomposition
InInt Conf Comput Vis pages 24882495 2013 8
10 Geonho Cha Minsik Lee Jungchan Cho and Songhwai Oh
Reconstruct as far as you can Consensus of nonrigid recon
struction from feasible regions IEEE Trans Pattern Anal
Mach Intell 432623637 2019 2
11 Yuchao Dai Hongdong Li and Mingyi He A simple prior
free method for nonrigid structurefrommotion factoriza
tion Int J Comput Vis 1072101122 2014 1 2 3 5
6 7 8
12 Alessio Del Bue A factorization approach to structure from
motion with shape priors In IEEE Conf Comput Vis Pat
tern Recog pages 18 2008 2
13 Alessio Del Bue Xavier Llad and Lourdes Agapito Non
rigid metric shape and motion recovery from uncalibrated
images using priors In IEEE Conf Comput Vis Pattern
Recog pages 11911198 2006 2 7
14 Hui Deng Tong Zhang Yuchao Dai Jiawei Shi Yiran
Zhong and Hongdong Li Deep nonrigid structurefrom
motion A sequencetosequence translation perspective
arXiv preprint arXiv220404730 2022 1
15 Paulo FU Gotardo and Aleix M Martinez Computing
smooth time trajectories for camera and deformable shape
in structure from motion with occlusion IEEE Trans Pattern Anal Mach Intell 331020512065 2011 1 2 7
9
16 Paulo FU Gotardo and Aleix M Martinez Kernel nonrigid
structure from motion In Int Conf Comput Vis page 802
2011 7
17 Paulo FU Gotardo and Aleix M Martinez Nonrigid struc
ture from motion with complementary rank3 spaces In
IEEE Conf Comput Vis Pattern Recog pages 30653072
2011 2 7 9
18 John C Gower Generalized procrustes analysis Psychome
trika 403351 1975 2
19 Stella Graßhof and Sami Sebastian Brandt Tensorbased
nonrigid structure from motion In IEEE Winter Conference
on Applications of Computer VisionWACV pages 3011
3020 2022 2
20 Mei Han and Takeo Kanade Reconstruction of a scene with
multiple linearly moving objects Int J Comput Vis 59
285300 2004 2
21 Jos e Pedro Iglesias Carl Olsson and Marcus Valto

12 The comparison results are reported in Tab 5 SWNN

improves the accuracy of 3D shape recovery in most de

formation settings The accuracy improvement mainly de

pends on the segmentation result and the smoothing prop

erty of the sequence for more analysis see Supp Sec 3

Performance on Noisy and Missing Data We follow the

approaches in 30 to add noise and occlusion to the data

For missing data we first complete the measurement ma

trix by 9 and then run our method Fig 5 shows that our

method performs better than others in most cases3

5 Conclusion

In this paper we have proposed a spatialtemporal modeling

framework for NRSfM We proposed a Temporallysmooth

Procrustean Alignment module to tackle the rotation ambi

guity Furthermore we introduced a spatialweighted nu

clear norm model with the shape proxy strategy to ensure

the effectiveness of lowrank constraint in coping with se

vere spatiallyvariant deformations Our method achieves

superior 3D reconstruction performance across a wide range

of deformation sequences Moreover there is still much re

search to do in the future on more accurate trajectory seg

mentation and explicit modeling of complex deformation

Acknowledgements This research was supported in part

by the National Natural Science Foundation of China

62271410 and the Fundamental Research Funds for the

Central Universities

3See Supp for more tests on TPA SWNN and Missing DataReferences

1 Antonio Agudo Unsupervised 3d reconstruction and group

ing of rigid and nonrigid categories IEEE Trans Pattern

Anal Mach Intell 441519532 2020 1 2

2 Ijaz Akhter Yaser Sheikh Sohaib Khan and Takeo Kanade

Nonrigid structure from motion in trajectory space In Adv

Neural Inform Process Syst pages 4148 2008 1 2 6 7

3 Ijaz Akhter Yaser Sheikh and Sohaib Khan In defense of

orthonormality constraints for nonrigid structure from mo

tion In IEEE Conf Comput Vis Pattern Recog pages

15341541 2009 2

4 Benedicte Bascle and Andrew Blake Separability of pose

and expression in facial tracking and animation In Int Conf

Comput Vis pages 323328 1998 2

5 V olker Blanz and Thomas Vetter A morphable model for

the synthesis of 3d faces In Proceedings of the 26th an

nual conference on Computer graphics and interactive tech

niques pages 187194 1999 2

6 Stephen Boyd Neal Parikh Eric Chu Borja Peleato

Jonathan Eckstein et al Distributed optimization and sta

tistical learning via the alternating direction method of mul

tipliers Foundations and Trends in Machine learning 3

11122 2011 6

7 Matthew Brand A direct method for 3d factorization of non

rigid motion observed in 2d In IEEE Conf Comput Vis

Pattern Recog pages 122128 2005 2

8 Christoph Bregler Aaron Hertzmann and Henning Bier

mann Recovering nonrigid 3d shape from image streams

InIEEE Conf Comput Vis Pattern Recog pages 690696

2000 1 2 3

9 Ricardo Cabral Fernando De la Torre Jo ao P Costeira and

Alexandre Bernardino Unifying nuclear norm and bilinear

factorization approaches for lowrank matrix decomposition

InInt Conf Comput Vis pages 24882495 2013 8

10 Geonho Cha Minsik Lee Jungchan Cho and Songhwai Oh

Reconstruct as far as you can Consensus of nonrigid recon

struction from feasible regions IEEE Trans Pattern Anal

Mach Intell 432623637 2019 2

11 Yuchao Dai Hongdong Li and Mingyi He A simple prior

free method for nonrigid structurefrommotion factoriza

tion Int J Comput Vis 1072101122 2014 1 2 3 5

6 7 8

12 Alessio Del Bue A factorization approach to structure from

motion with shape priors In IEEE Conf Comput Vis Pat

tern Recog pages 18 2008 2

13 Alessio Del Bue Xavier Llad and Lourdes Agapito Non

rigid metric shape and motion recovery from uncalibrated

images using priors In IEEE Conf Comput Vis Pattern

Recog pages 11911198 2006 2 7

14 Hui Deng Tong Zhang Yuchao Dai Jiawei Shi Yiran

Zhong and Hongdong Li Deep nonrigid structurefrom

motion A sequencetosequence translation perspective

arXiv preprint arXiv220404730 2022 1

15 Paulo FU Gotardo and Aleix M Martinez Computing

smooth time trajectories for camera and deformable shape

in structure from motion with occlusion IEEE Trans Pattern Anal Mach Intell 331020512065 2011 1 2 7

9

16 Paulo FU Gotardo and Aleix M Martinez Kernel nonrigid

structure from motion In Int Conf Comput Vis page 802

2011 7

17 Paulo FU Gotardo and Aleix M Martinez Nonrigid struc

ture from motion with complementary rank3 spaces In

IEEE Conf Comput Vis Pattern Recog pages 30653072

2011 2 7 9

18 John C Gower Generalized procrustes analysis Psychome

trika 403351 1975 2

19 Stella Graßhof and Sami Sebastian Brandt Tensorbased

nonrigid structure from motion In IEEE Winter Conference

on Applications of Computer VisionWACV pages 3011

3020 2022 2

20 Mei Han and Takeo Kanade Reconstruction of a scene with

multiple linearly moving objects Int J Comput Vis 59

285300 2004 2

21 Jos e Pedro Iglesias Carl Olsson and Marcus Valto