MIT Saliency Benchmark

NEWS:

The goal of this website is to be the most up-to-date, online source of saliency model performances and datasets. We believe that a continuously updated all-in-one comparison page will serve as an essential resource to document and promote progress in the field of saliency modeling.

On this site, our contribution is twofold:
(1) We score and report performances for the latest saliency models on our saliency benchmark: the only data sets where human eye movements are not public, in order to prevent training on, and fitting to, the datasets. Go to model performances or submit a new model.
(2) We maintain an up-to-date listing of other saliency datasets and collection procedures, along with all relevant links in order to have a one-stop resource and thereby to facilitate comparison. Go to datasets.

website citation

@misc{mit-saliency-benchmark,
  author       = {Zoya Bylinskii and Tilke Judd and Ali Borji and Laurent Itti and Fr{\'e}do Durand and Aude Oliva and Antonio Torralba},
  title        = {MIT Saliency Benchmark},
  howpublished = {http://saliency.mit.edu/}
}

paper citations

@article{salMetrics_Bylinskii,
    title    = {What do different evaluation metrics tell us about saliency models?},
    author   = {Zoya Bylinskii and Tilke Judd and Aude Oliva and Antonio Torralba and Fr{\'e}do Durand},
    journal  = {arXiv preprint arXiv:1604.03605},
    year     = {2016}
}

@InProceedings{Judd_2012,
   title     = {A Benchmark of Computational Models of Saliency to Predict Human Fixations},
   author    = {Tilke Judd and Fr{\'e}do Durand and Antonio Torralba},
   booktitle = {MIT Technical Report},
   year      = {2012}
}

@article{borji2013quantitative,
   title     = {Quantitative analysis of human-model agreement in visual saliency modeling: A comparative study},
   author    = {Borji, Ali and Sihite, Dicky N and Itti, Laurent},
   journal   = {Image Processing, IEEE Transactions on},
   volume    = {22},
   number    = {1},
   pages     = {55--69},
   year      = {2013},
   publisher = {IEEE}
}

@article{CAT2000,
   title     = {CAT2000: A Large Scale Fixation Dataset for Boosting Saliency Research},
   author    = {Borji, Ali and Itti, Laurent},
   journal   = {CVPR 2015 workshop on "Future of Datasets"},
   year      = {2015},
   note      = {arXiv preprint arXiv:1505.03581}
}