作者: Eric Saund , Prateek Sarkar , Jing Lin
DOI:
关键词: Artificial intelligence 、 Mode (computer interface) 、 Pixel 、 State (computer science) 、 Executable 、 Computer vision 、 Image (mathematics) 、 Recognition algorithm 、 User interface 、 Computer science
摘要: A user interface and method is embodied on a computer readable medium executable computer. The labeler which labels only foreground pixels of an image stored in computing environment. operates Region mode/state Brush mode/state, includes Tentative mode that permits assigned label to be changed after have been selected. Groups may selected for labeling at once by point-and-click command, pixel belong one or more groups are memory as layers. formed dynamically selection actions, and/or through automatic recognition algorithms. Pixels already labeled with certain locked not altered additional operations. Unassigned highlighted increase the ease they identified image. Comparisons between images undertaken indicate differences different groundtruth labeling.