Consistent Saliency Benchmarking: How One Model Can Win on All Metrics

Authors

Matthias Kümmerer, Centre for Integrative Neuroscience, TübingenFollow
Thomas S.A. Wallis, Centre for Integrative Neuroscience, Tübingen
Matthias Bethge, University of Tuebingen

Keywords

Saliency, Fixations, Benchmarking, Model-Comparison, Metrics, Bayesian Utility Theory

Start Date

16-5-2018 9:50 AM

End Date

16-5-2018 10:15 AM

Recommended Citation

Kümmerer, Matthias; Wallis, Thomas S.A.; and Bethge, Matthias, "Consistent Saliency Benchmarking: How One Model Can Win on All Metrics" (2018). MODVIS Workshop. 2.
https://docs.lib.purdue.edu/modvis/2018/session01/2