view article
Figure 2
The proposed CNN for screening X-ray images consists of four sets of convolution / batch normalization / rectification (ReLU) / downsampling (MaxPooling) layers, as detailed in §2.1[link]. (a) Numbers and the sizes of the output feature response maps produced by each of the filter layers. The resulting feature map is concatenated (flattened) into a one-dimensional feature vector, and fed into the feature classification stage. (b) Details of the particular sequence of filters that is responsible for the transformation from the 4 × 180 × 180 image stack to the 8 × 45 × 45 stack.

Journal logoJOURNAL OF
SYNCHROTRON
RADIATION
ISSN: 1600-5775
Volume 25| Part 3| May 2018| Pages 655-670
https://doi.org/10.1107/S1600577518004873
Follow J. Synchrotron Rad.
Sign up for e-alerts
E-alerts
Follow J. Synchrotron Rad. on Twitter
Twitter
Follow us on facebook
Facebook
Sign up for RSS feeds
RSS