Skip to main content
. 2019 Nov 18;62(3):417–444. doi: 10.1007/s10851-019-00919-7

Fig. 6.

Fig. 6

Image inpainting from incomplete data. From left to right: Data, TGV-based reconstruction, proposed method (only TXT in first row), nine most important learned atoms. Rows 1, 3: 20% of pixels, rows 2, 4: 30% of pixels