• Catherine Qi Zhao

    Assistant Professor
    Department of Computer Science and Engineering
    University of Minnesota

    Office: Keller 6-189
    Phone: (612) 301-2115
    Email: qzhao at cs.umn.edu

    Adjunct Assistant Professor
    Department of Electrical and Computer Engineering
    National University of Singapore

  • Our research focuses on providing theoretical foundations and computational innovations in the study of computational and human vision. We take an integrated experimental and computational approach with theories and tools from computer vision, machine learning, visual cognition, computational neuroscience, and physics to develop quantitative models of biological and computational vision, to inspire interesting experimental design for psychophysical and neurophysiological study, to gain insights into visual and cognitive disorders, and to build intelligent visual systems that approach human performance.

    Questions we ask include: how selective attention is represented and processed in the human brain? and how is it for computers? Can we bridge the two despite the fundamentally different processing units and mechanisms of the brain and the computer? Most of the work in our lab studies these problems with complex natural scenes.



  • Database

    • SALICON database. Saliency in Context - a large-scale attention database on MS COCO images. Jiang et al. CVPR [pdf] [bib]
    • OSIE database. Object and Semantic Images and Eye-tracking database - a database for object and semantic saliency (700 images, 5551 objects with fine contour and semantic attribute labeling). Xu et al. JoV [pdf] [bib]
    • EyeCrowd database. Eye Fixations in Crowd database - a database for saliency in crowd. Jiang et al. ECCV [pdf] [bib]


  • Code

    • Saliency with Objects and Semantics. Code for object and semantic feature computation, model training with SVM, saliency prediction, and evaluation measures.
    • Saliency in Crowd. Code for crowd feature computation, crowd stats calculation, model training with MKL, saliency prediction, and evaluations.
    • Multi-Layer Sparse Network. Code for multi-layer sparse network, model training, and saliency prediction.