Click on a metric name to sort that metric. Models evaluated as probabilistic models are shown with green background. The performance under the metric which a model has been trained on is shown in italic. The code for evaluation models can be found here.
Name  Published  Code  IG  AUC  sAUC  NSS  CC  KLDiv  SIM  Date tested 
Saliency Detection by SelfResemblance (SSR) 
Hae Jong Seo, Peyman Milanfar Nonparametric BottomUp Saliency Detection by SelfResemblance [CVPR 2009]. 
0.7078  0.6184  0.7851  0.3007  1.6363  0.3832 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

Lingyun Zhang, Matthew H. Tong, Tim K. Marks, Honghao Shan, Garrison W. Cottrell. SUN: A Bayesian framework for saliency using natural statistics [JoV 2008] 
0.7270  0.6358  0.8709  0.3129  1.3096  0.3713 
First tested 20230715 Last tested 20230715 maps from code via SMILER. Params: rescale=0.5 

QuaternionBased Spectral Saliency (QSS) 
B. Schauerte, R. Stiefelhagen. Quaternionbased Spectral Saliency Detection for Eye Fixation Prediction [ECCV 2012] 
0.7369  0.6343  0.9272  0.3504  1.2178  0.3933 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

Image Signature 
Xiaodi Hou, Jonathan Harel, Christof Koch. Image Signature: Highlighting Sparse Salient Regions [PAMI 2011] 
0.7543  0.6271  0.9822  0.3778  1.1881  0.3935 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

Stas Goferman, Lihi ZelnikManor, Ayellet Tal. ContextAware Saliency Detection [CVPR 2010] [PAMI 2012] 
0.7616  0.6179  0.9871  0.3783  1.1575  0.3987 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

Dynamic Visual Attention (DVA) 
Hou, Xiaodi, and Liqing Zhang. Dynamic visual attention: Searching for coding length increments NIPS 2008 
0.7686  0.6448  1.0503  0.3916  1.1569  0.4309 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

AIM 
Neil Bruce, John Tsotsos. Attention based on information maximization [JoV 2007] 
0.7729  0.6631  1.0010  0.3747  1.2724  0.3749 
First tested 20230714 Last tested 20230714 maps from code via SMILER. Params: resize=0.5, thebasis='31infomax975' 

Nicolas Riche, Matei Mancas, Matthieu Duvinage, Makiese Mibulumukini, Bernard Gosselin, Thierry Dutoit. RARE2012: A multiscale raritybased saliency detection with its comparative statistical analysis [Signal Processing: Image Communication, 2013] 
0.7789  0.6482  1.1460  0.4284  1.0822  0.4275 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

IttiKoch2 
Implementation by Jonathan Harel (part of GBVS toolbox) 
0.7810  0.6077  1.0939  0.4248  1.0838  0.4232 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

Jianming Zhang, Stan Sclaroff. Saliency detection: a boolean map approach [ICCV 2013] 
0.7978  0.6773  1.2373  0.4558  1.0304  0.4325 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

Hamed Rezazadegan Tavakoli, Esa Rahtu, Janne Heikkila. Fast and efficient saliency detection using sparse sampling and kernel density estimation [SCIA 2011] 
0.7978  0.5805  1.2224  0.4622  1.9651  0.4734 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

Jonathan Harel, Christof Koch, Pietro Perona. GraphBased Visual Saliency [NIPS 2006] 
0.8003  0.6021  1.2053  0.4609  1.0158  0.4404 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

Centerbias 
kernel density from training fixations with uniform mixture component, hyperparameters estimated on training dataset with leaveoneimageout crossvalidation 
0.0000  0.8008  0.4999  1.1979  0.4768  0.9724  0.4838 
First tested 20230704 Last tested 20230704 

Eleonora Vig, Michael Dorr, David Cox. LargeScale Optimization of Hierarchical Features for Saliency Prediction in Natural Images [CVPR 2014] 
0.8191  0.5953  1.1260  0.4438  1.2378  0.3713 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

Erkut Erdem, Aykut Erdem. Visual saliency estimation by nonlinearly integrating features using region covariances [JoV 2013] 
0.8197  0.5705  1.3061  0.4956  1.5682  0.4938 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

LDS 
Shu Fang, Jia Li, Yonghong Tian, Tiejun Huang, Xiaowu Chen. Learning Discriminative Subspaces on Random Contrasts for Image Saliency Analysis [TNNLS 2016] 
0.8213  0.5931  1.3996  0.5310  0.9695  0.5098 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

MLNet 
Marcella Cornia, Lorenzo Baraldi, Giuseppe Serra, Rita Cucchiara. A Deep MultiLevel Network for Saliency Prediction [ICPR 2016] 
0.8335  0.7122  1.6734  0.5895  0.8382  0.5128 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

Gold Standard (leaveonesubjectout) 
Gaussian kernel density estimate using all fixations of an image with uniform mixture component. Crossvalidated over subjects. Leaveonesubjectout performance 
0.8339  0.7667  1.3976 
First tested 20240207 Last tested 20240207 

Saliency Attentive Model (SAMVGG) 
Marcella Cornia, Lorenzo Baraldi, Giuseppe Serra, Rita Cucchiara. Predicting Human Eye Fixations via an LSTMbased Saliency Attentive Model [IEEE TIP 2018] 
0.8430  0.6947  1.6829  0.5974  1.0533  0.5453 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

DeepGaze I 
Matthias Kümmerer, Lucas Theis, Matthias Bethge. Deep Gaze I: Boosting Saliency Prediction with Feature Maps Trained on ImageNet [arxiv 2014, ICLR 2015 workshop] 
0.2252  0.8436  0.6891  1.5509  0.5674  0.8331  0.5278 
First tested 20230721 Last tested 20230721 predictions from authors 

EMLNET 
Sen Jia & Neil D.B. Bruce EMLNET: An Expandable MultiLayer NETwork for Saliency Prediction [arXiv 2018] 
0.8447  0.7069  1.7354  0.6067  1.3571  0.5512 
First tested 20240514 Last tested 20240514 maps computed from model source code using a downscaling which was found to be optimal on the training set 

Saliency Attentive Model (SAMResNet) 
Marcella Cornia, Lorenzo Baraldi, Giuseppe Serra, Rita Cucchiara. Predicting Human Eye Fixations via an LSTMbased Saliency Attentive Model [IEEE TIP 2018] 
0.8478  0.6985  1.7695  0.6261  1.0553  0.5612 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

Junting Pan, Cristian Canton, Kevin McGuinness, Noel E. O'Connor, Jordi Torres, Elisa Sayrol and Xavier GiroiNieto. SalGAN: Visual Saliency Prediction with Generative Adversarial Networks [arXiv 2017] 
0.8500  0.7019  1.6702  0.6227  0.8117  0.5495 
First tested 20230713 Last tested 20230713 saliency maps computed with SMILER 

X. Hun C. Shen, X. Boix, Q. Zhao: SALICON: Reducing the Semantic Gap in Saliency Prediction by Adapting Deep Neural Networks. ICCV 2015 
0.8640  0.7188  1.8301  0.6690  0.6626  0.5845 
First tested 20230724 Last tested 20230724 maps from SALICON web demo 

Matthias Kümmerer, Thomas S. A. Wallis, Leon A. Gatys, Matthias Bethge. Understanding Low and HighLevel Contributions to Fixation Prediction [ICCV 2017] 
0.6636  0.8699  0.7399  2.0028  0.6909  0.5858  0.6043 
First tested 20230725 Last tested 20230725 

G. Ding, N. Imamoglu, A. Caglayan, M. Murakawa, R. Nakamura: SalFBNet: Learning PseudoSaliency Distribution via Feedback Convolutional Networks. Image and Vision Computing 2022. 
0.8722  0.7099  2.0275  0.7088  0.8623  0.6178 
First tested 20240514 Last tested 20240514 saliency maps computed with public code 

R. Droste, J. Jiao, J.A. Noble: Unified Image and Video Saliency Modeling. ECCV 2020 (arXiv) 
0.7494  0.8774  0.7585  2.0954  0.7155  0.5515  0.6203 
First tested 20240216 Last tested 20240216 predictions from public model implementation. Images have been downscaled by a factor, which was found to optimal on the COCO Freeview training set 

A. Linardos, M. Kümmerer, O. Press, M. Bethge: DeepGaze IIE: Calibrated prediction in and outofdomain for stateoftheart saliency modeling [ICCV 2021] 
0.8596  0.8825  0.7669  2.2558  0.7563  0.4863  0.6447 
First tested 20230721 Last tested 20230721 densities from authors 

Gold Standard (leaveonesubjectout) 
Gaussian kernel density estimate using all fixations of an image with uniform mixture component. Crossvalidated over subjects. Leaveonesubjectout performance 
0.8673  0.8829  2.2837 
First tested 20240206 Last tested 20240206 