Fig. 2: Performance benchmarks for ELD.
From: Spatial landmark detection and tissue registration with deep learning
a–d, The performance benchmarks for ELD and two other state-of-the-art models for landmark detection. Benchmarks were conducted on the MAFL and AFLW datasets with 1,000 and 2,991 samples, respectively. The parts show the distribution of the mean consistency error for each image (a), consistency error for each landmark (b), and backward (c) and forward errors (d) for each image. The box plots display the median (central line within each box), the interquartile range (boxes representing the 25th to 75th percentiles) and the whiskers extend to 1.5 times the interquartile range above and below the box.
