SuperSlicing Frame Restoration for Anisotropic ssTEM and Video Data

Dmitry Laptev, Joachim M. Buhmann
Proceedings of the Neural Connectomics Workshop at ECML 2014, PMLR 46:93-103, 2015.

Abstract

In biological imaging the data is often represented by a sequence of anisotropic frames — the resolution in one dimension is significantly lower than in the other dimensions. E.g. in electron microscopy it arises from the thickness of a scanned section. This leads to blurred images and raises problems in tasks like neuronal image segmentation. We present the details and additional evaluation of an approach originally introduced in Laptev et al. (2014) called SuperSlicing to decompose the observed frame into a sequence of plausible hidden sub-frames. Based on sub-frame decomposition by SuperSlicing we propose a novel automated method to perform neuronal structure segmentation. We test our approach on a popular connectomics benchmark, where SuperSlicing preserves topological structures significantly better than other algorithms. We also generalize the approach for video anisotropicity that comes from the long exposure time and show that our method outperforms baseline methods on a reconstruction of low frame rate videos of natural scenes.

Cite this Paper

BibTeX
@InProceedings{pmlr-v46-laptev15, title = {SuperSlicing Frame Restoration for Anisotropic ssTEM and Video Data}, author = {Laptev, Dmitry and Buhmann, Joachim M.}, booktitle = {Proceedings of the Neural Connectomics Workshop at ECML 2014}, pages = {93--103}, year = {2015}, editor = {Battaglia, Demian and Guyon, Isabelle and Lemaire, Vincent and Soriano, Jordi}, volume = {46}, series = {Proceedings of Machine Learning Research}, month = {15 Sep}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v46/laptev15.pdf}, url = {https://proceedings.mlr.press/v46/laptev15.html}, abstract = {In biological imaging the data is often represented by a sequence of anisotropic frames — the resolution in one dimension is significantly lower than in the other dimensions. E.g. in electron microscopy it arises from the thickness of a scanned section. This leads to blurred images and raises problems in tasks like neuronal image segmentation. We present the details and additional evaluation of an approach originally introduced in Laptev et al. (2014) called SuperSlicing to decompose the observed frame into a sequence of plausible hidden sub-frames. Based on sub-frame decomposition by SuperSlicing we propose a novel automated method to perform neuronal structure segmentation. We test our approach on a popular connectomics benchmark, where SuperSlicing preserves topological structures significantly better than other algorithms. We also generalize the approach for video anisotropicity that comes from the long exposure time and show that our method outperforms baseline methods on a reconstruction of low frame rate videos of natural scenes. } }
Endnote
%0 Conference Paper %T SuperSlicing Frame Restoration for Anisotropic ssTEM and Video Data %A Dmitry Laptev %A Joachim M. Buhmann %B Proceedings of the Neural Connectomics Workshop at ECML 2014 %C Proceedings of Machine Learning Research %D 2015 %E Demian Battaglia %E Isabelle Guyon %E Vincent Lemaire %E Jordi Soriano %F pmlr-v46-laptev15 %I PMLR %P 93--103 %U https://proceedings.mlr.press/v46/laptev15.html %V 46 %X In biological imaging the data is often represented by a sequence of anisotropic frames — the resolution in one dimension is significantly lower than in the other dimensions. E.g. in electron microscopy it arises from the thickness of a scanned section. This leads to blurred images and raises problems in tasks like neuronal image segmentation. We present the details and additional evaluation of an approach originally introduced in Laptev et al. (2014) called SuperSlicing to decompose the observed frame into a sequence of plausible hidden sub-frames. Based on sub-frame decomposition by SuperSlicing we propose a novel automated method to perform neuronal structure segmentation. We test our approach on a popular connectomics benchmark, where SuperSlicing preserves topological structures significantly better than other algorithms. We also generalize the approach for video anisotropicity that comes from the long exposure time and show that our method outperforms baseline methods on a reconstruction of low frame rate videos of natural scenes.
RIS
TY - CPAPER TI - SuperSlicing Frame Restoration for Anisotropic ssTEM and Video Data AU - Dmitry Laptev AU - Joachim M. Buhmann BT - Proceedings of the Neural Connectomics Workshop at ECML 2014 DA - 2015/10/21 ED - Demian Battaglia ED - Isabelle Guyon ED - Vincent Lemaire ED - Jordi Soriano ID - pmlr-v46-laptev15 PB - PMLR DP - Proceedings of Machine Learning Research VL - 46 SP - 93 EP - 103 L1 - http://proceedings.mlr.press/v46/laptev15.pdf UR - https://proceedings.mlr.press/v46/laptev15.html AB - In biological imaging the data is often represented by a sequence of anisotropic frames — the resolution in one dimension is significantly lower than in the other dimensions. E.g. in electron microscopy it arises from the thickness of a scanned section. This leads to blurred images and raises problems in tasks like neuronal image segmentation. We present the details and additional evaluation of an approach originally introduced in Laptev et al. (2014) called SuperSlicing to decompose the observed frame into a sequence of plausible hidden sub-frames. Based on sub-frame decomposition by SuperSlicing we propose a novel automated method to perform neuronal structure segmentation. We test our approach on a popular connectomics benchmark, where SuperSlicing preserves topological structures significantly better than other algorithms. We also generalize the approach for video anisotropicity that comes from the long exposure time and show that our method outperforms baseline methods on a reconstruction of low frame rate videos of natural scenes. ER -
APA
Laptev, D. & Buhmann, J.M.. (2015). SuperSlicing Frame Restoration for Anisotropic ssTEM and Video Data. Proceedings of the Neural Connectomics Workshop at ECML 2014, in Proceedings of Machine Learning Research 46:93-103 Available from https://proceedings.mlr.press/v46/laptev15.html.

Related Material