Publication

Efficient convolutional neural networks for pixelwise classification on heterogeneous hardware systems

Conference Article

Conference

IEEE International Symposium on Biomedical Imaging (ISBI)

Edition

13th

Pages

1225-1228

Doc link

http://dx.doi.org/10.1109/ISBI.2016.7493487

File

Download the digital copy of the doc pdf document

Authors

  • Tschopp, Fabian

  • Martel, Julien N.P.

  • Turaga, Srinivas C.

  • Cook, Matthew

  • Funke, Jan

Abstract

With recent advances in high-throughput Electron Microscopy (EM) imaging it is now possible to image an entire nervous system of organisms like Drosophila melanogaster. One of the bottlenecks to reconstruct a connectome from these large volumes (≈ 100 TiB) is the pixel-wise prediction of membranes. The time it would typically take to process such a volume using a convolutional neural network (CNN) with a sliding window approach is in the order of years on a current GPU. With sliding windows, however, a lot of redundant computations are carried out. In this paper, we present an extension to the Caffe library to increase throughput by predicting many pixels at once. On a sliding window network successfully used for membrane classification, we show that our method achieves a speedup of up to 57×, maintaining identical prediction results.

Categories

computer vision.

Author keywords

semantic segmentation, convolutional neural network, electron microscopy

Scientific reference

F. Tschopp, J.N. Martel, S.C. Turaga, M. Cook and J. Funke. Efficient convolutional neural networks for pixelwise classification on heterogeneous hardware systems, 13th IEEE International Symposium on Biomedical Imaging, 2016, Prague, pp. 1225-1228.