Meta Learning with Differentiable Closed-form Solver for Fast Video Object Segmentation

    Cited by: 1|Bibtex|Views12|Links

    Abstract:

    This paper tackles the problem of video object segmentation. We are specifically concerned with the task of segmenting all pixels of a target object in all frames, given the annotation mask in the first frame. Even when such annotation is available this remains a challenging problem because of the changing appearance and shape of the ob...More

    Code:

    Data:

    Your rating :
    0

     

    Tags
    Comments