Quality control in crowdsourced object segmentation

This paper explores processing techniques to deal with noisy data in crowdsourced object segmentation tasks. We use the data collected with Click´n´Cut, an online interactive segmentation tool, and we perform several experiments towards improving the segmentation results. First, we introduce different superpixel-based techniques to filter users´ traces, and assess their impact on the segmentation result. Second, we present different criteria to detect and discard the traces from potential bad users, resulting in a remarkable increase in performance. Finally, we show a novel superpixel-based segmentation algorithm which does not require any prior filtering and is based on weighting each user´s contribution according to his/her level of expertise.