Skip to main content
. 2022 May 10;17(5):e0267452. doi: 10.1371/journal.pone.0267452

Fig 3. Realistic labelled single molecule data are synthesised by DeepGANnel.

Fig 3

(A) Discriminator and Generator losses plotted per epoch during training. (B) The maximum mean discrepancy (MMD) and dynamic time warping (DTW) evolution over the first 250 epochs as they approach equilibrium. (C and D) Examples of real (C) and GAN synthesised (D) labelled data records. (E and F) All points amplitude histograms calculated from the real (E) and GAN synthesised (F) data.