TY - GEN
T1 - RISEC
T2 - 14th International Symposium on Visual Computing, ISVC 2019
AU - Memariani, Ali
AU - Endres, Bradley T.
AU - Bassères, Eugénie
AU - Garey, Kevin W.
AU - Kakadiaris, Ioannis A.
N1 - Funding Information:
This work was supported in part by NIH/NIAID 1UO1 AI-24290-01 and by the Hugh Roy and Lillie Cranz Cullen Endowment Fund. At the time of data collection. Dr. Endres was a postdoctoral fellow at the University of Houston. All statements of facts, opinion or conclusions contained herein are those of the authors and should not be construed as representing official views or policies of the sponsors.
Funding Information:
Acknowledgments. This work was supported in part by NIH/NIAID 1UO1 AI-24290-01 and by the Hugh Roy and Lillie Cranz Cullen Endowment Fund. At the time of data collection. Dr. Endres was a postdoctoral fellow at the University of Houston. All statements of facts, opinion or conclusions contained herein are those of the authors and should not be construed as representing official views or policies of the sponsors.
Publisher Copyright:
© 2019, Springer Nature Switzerland AG.
PY - 2019
Y1 - 2019
N2 - Detection of Clostridioides difficile cells in scanning electron microscopy images is a challenging task due to the challenges of cell rotation and inhomogeneous illumination. Currently, orientation-invariance in deep ConvNets is achieved by data augmentation. However, training with all possible orientations increases computational complexity. Furthermore, conventional illumination-invariance models include pre-processing illumination normalization steps. However, illumination normalization algorithms remove important texture information which is critical for the analysis of SEM images. In this paper, RISEC (Rotational Invariant Segmentation of Elongated Cells in SEM images with Inhomogeneous Illumination) is proposed to address the challenges of cell rotation and inhomogeneous illumination. First, a generative adversarial network segments the candidate cell regions proposals, addressing the inhomogeneous illumination. Then, the region proposals are passed to two capsule layers where a rotation-invariant shape representation is learned for every cell type via dynamic routing. Our experiments indicate that RISEC outperforms the state of the art models (e.g., CapsNet, and U-net) by at least 11% improving the dice score.
AB - Detection of Clostridioides difficile cells in scanning electron microscopy images is a challenging task due to the challenges of cell rotation and inhomogeneous illumination. Currently, orientation-invariance in deep ConvNets is achieved by data augmentation. However, training with all possible orientations increases computational complexity. Furthermore, conventional illumination-invariance models include pre-processing illumination normalization steps. However, illumination normalization algorithms remove important texture information which is critical for the analysis of SEM images. In this paper, RISEC (Rotational Invariant Segmentation of Elongated Cells in SEM images with Inhomogeneous Illumination) is proposed to address the challenges of cell rotation and inhomogeneous illumination. First, a generative adversarial network segments the candidate cell regions proposals, addressing the inhomogeneous illumination. Then, the region proposals are passed to two capsule layers where a rotation-invariant shape representation is learned for every cell type via dynamic routing. Our experiments indicate that RISEC outperforms the state of the art models (e.g., CapsNet, and U-net) by at least 11% improving the dice score.
KW - Convolutional capsules
KW - Generative adversarial networks
KW - Illumination
KW - Instance segmentation
KW - Orientation-invariance
UR - http://www.scopus.com/inward/record.url?scp=85076182443&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85076182443&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-33723-0_45
DO - 10.1007/978-3-030-33723-0_45
M3 - Conference contribution
AN - SCOPUS:85076182443
SN - 9783030337223
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 553
EP - 563
BT - Advances in Visual Computing - 14th International Symposium on Visual Computing, ISVC 2019, Proceedings
A2 - Bebis, George
A2 - Parvin, Bahram
A2 - Boyle, Richard
A2 - Koracin, Darko
A2 - Ushizima, Daniela
A2 - Chai, Sek
A2 - Sueda, Shinjiro
A2 - Lin, Xin
A2 - Lu, Aidong
A2 - Thalmann, Daniel
A2 - Wang, Chaoli
A2 - Xu, Panpan
PB - Springer
Y2 - 7 October 2019 through 9 October 2019
ER -