Skip to main content
Figure 6 | Fixed Point Theory and Algorithms for Sciences and Engineering

Figure 6

From: Learning without loss

Figure 6

Left: Sixteen sample images from the letter-montage data set. Right: The 16 features obtained in one run of RRR. Letters in the data occur with two fonts (plain and slant); in a successful NMF all the recovered features/letters are weakly slanted

Back to article page