Share Email Print

Proceedings Paper

Trade-off between speed and performance for colorectal endoscopic NBI image classification
Author(s): Shoji Sonoyama; Toru Tamaki; Tsubasa Hirakawa; Bisser Raytchev; Kazufumi Kaneda; Tetsushi Koide; Yoko Kominami; Shigeto Yoshida; Shinji Tanaka
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

This paper investigates a trade-off between computation time and recognition rate of local descriptor-based recognition for colorectal endoscopic NBI image classification. Recent recognition methods using descriptors have been successfully applied to medical image classification. The accuracy of these methods might depend on the quality of vector quantization (VQ) and encoding of descriptors, however an accurate quantization takes a long time. This paper reports how a simple sampling strategy affects performances with different encoding methods. First, we extract about 7.7 million local descriptors from training images of a dataset of 908 NBI endoscopic images. Second, we randomly choose a subset of between 7.7M and 19K descriptors for VQ. Third, we use three encoding methods (BoVW, VLAD, and Fisher vector) with different number of descriptors. Linear SVM is used for classification of a three-class problem. The computation time for VQ was drastically reduced by the factor of 100, while the peak performance was retained. The performance improved roughly 1% to 2% when more descriptors by over-sampling were used for encoding. Performances with descriptors extracted every pixel ("grid1") or every two pixels ("grid2") are similar, while the computation time is very different; grid2 is 5 to 30 times faster than grid1. The main finding of this work is twofold. First, recent encoding methods such as VLAD and Fisher vector are as insensitive to the quality of VQ as BoVW. Second, there is a trade-off between computation time and performance in encoding over-sampled descriptors with BoVW and Fisher vector, but not with VLAD.

Paper Details

Date Published: 20 March 2015
PDF: 8 pages
Proc. SPIE 9413, Medical Imaging 2015: Image Processing, 94132D (20 March 2015); doi: 10.1117/12.2081928
Show Author Affiliations
Shoji Sonoyama, Hiroshima Univ. (Japan)
Toru Tamaki, Hiroshima Univ. (Japan)
Tsubasa Hirakawa, Hiroshima Univ. (Japan)
Bisser Raytchev, Hiroshima Univ. (Japan)
Kazufumi Kaneda, Hiroshima Univ. (Japan)
Tetsushi Koide, Hiroshima Univ. (Japan)
Yoko Kominami, Hiroshima Univ. (Japan)
Shigeto Yoshida, Hiroshima Univ. (Japan)
Shinji Tanaka, Hiroshima Univ. (Japan)

Published in SPIE Proceedings Vol. 9413:
Medical Imaging 2015: Image Processing
Sébastien Ourselin; Martin A. Styner, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?