AIVE can also function in 2D. (A–E) Raw membrane predictions and (C and D) AIVE results for (E) one TEM image, made using six different machine learning algorithms; RF, J48, MLP (multilayer perceptron), DT (decision table), JRip, and PART (projective adaptive resonance theory). (B and D) Visual depiction of the absolute difference in voxel values (ΔV) between each result from A and C, showing the average value difference for each two-way value comparison as a percentage of the total dynamic range (0–1). (F) Quantification of the average ΔV in the images for each two-way comparison between each of the (A) model predictions and (B) each of the AIVE outputs generated using those same six models. Scale bars; A–E, 200 nm.