Abstract: We present a novel self-distillation based self-supervised monocular depth estimation (SD-SSMDE) learning framework. In the first step, our network is trained in a self-supervised regime on ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results