Fig. 7From: Coarse-to-fine online learning for hand segmentation in egocentric videoa The original image has similar color in background and hand regions. From (b–d) are corresponding results of ours, Li and Kitani [1] and Zhu et al. [2]Back to article page