Fig. 4: Classification of a Raman map for varying threshold values.
From: High-speed identification of suspended carbon nanotubes using Raman spectroscopy and deep learning
a Raman intensity map of the Silicon peak. b Raman intensity map of the G-peak. c Ground truth map. d Map of the predicted classes for varying softmax threshold values. e Occurrence of true positives, false positives, and false negatives for varying softmax threshold values for M-CNTs and S-CNTs. f Ratio of false-positive/true positives for varying softmax threshold values for M-CNTs and S-CNTs.
