Skip to main content
. Author manuscript; available in PMC: 2013 Feb 6.
Published in final edited form as: Med Image Comput Comput Assist Interv. 2012;15(Pt 3):345–352. doi: 10.1007/978-3-642-33454-2_43

Fig. 5. Slide-Level Performance.

Fig. 5

Left top: Comparison of our IKSVM-based algorithm’s performance to that of humans and the baseline method (GMM approach). Average Precision (AP) and maximum F1-measure across four test sets. Right: Slide-level Recall-Precision curves across different methods for one test set. Left bottom: Our algorithm’s slide-level performance for different object-level classifiers. Average Precision (AP) and Average Specificity (AS), where we average over four test sets.