Translation-invariant scene grouping | IEEE Conference Publication | IEEE Xplore