2019/09/01 by Mario Amrehn, Stefan Steidl, Amrehn, Mario +11 · 1 citation
Computer Science · Engineering · #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #FOS: Electrical engineering #Human-Computer Interaction (cs.HC) #Image and Video Processing (eess.IV) #Image and Video Quality Assessment #Industrial Vision Systems and Defect Detection #Machine Learning (cs.LG) #Visual Attention and Saliency Detection #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1909.00482
openalex publication_date 2019/09/01 · openalex created_date 2022/07/28 · openalex updated_date 2026/07/28
For complex segmentation tasks, the achievable accuracy of fully automated\nsystems is inherently limited. Specifically, when a precise segmentation result\nis desired for a small amount of given data sets, semi-automatic methods\nexhibit a clear benefit for the user. The optimization of human computer\ninteraction (HCI) is an essential part of interactive image segmentation.\nNevertheless, publications introducing novel interactive segmentation systems\n(ISS) often lack an objective comparison of HCI aspects. It is demonstrated,\nthat even when the underlying segmentation algorithm is the same throughout\ninteractive prototypes, their user experience may vary substantially. As a\nresult, users prefer simple interfaces as well as a considerable degree of\nfreedom to control each iterative step of the segmentation. In this article, an\nobjective method for the comparison of ISS is proposed, based on extensive user\nstudies. A summative qualitative content analysis is conducted via abstraction\nof visual and verbal feedback given by the participants. A direct assessment of\nthe segmentation system is executed by the users via the system usability scale\n(SUS) and AttrakDiff-2 questionnaires. Furthermore, an approximation of the\nfindings regarding usability aspects in those studies is introduced, conducted\nsolely from the system-measurable user actions during their usage of\ninteractive segmentation prototypes. The prediction of all questionnaire\nresults has an average relative error of 8.9%, which is close to the expected\nprecision of the questionnaire results themselves. This automated evaluation\nscheme may significantly reduce the resources necessary to investigate each\nvariation of a prototype's user interface (UI) features and segmentation\nmethodologies.\n