Abstract | ||
---|---|---|
In this paper, we present a novel approach to reconstruct missing areas in facial images by using a series of Restricted Boltzman Machines (RBMs). RBMs created with a low number of hidden neurons generalize well and are able to reconstruct basic structures in the missing areas. On the other hand networks with many hidden neurons tend to emphasize details, when using the reconstruction of the previous, more generalized RBMs, as their input. Since trained RBMs are fast in encoding and decoding data by design, our method is also suitable for processing video streams. |
Year | DOI | Venue |
---|---|---|
2015 | 10.1109/ISM.2015.68 | 2015 IEEE International Symposium on Multimedia (ISM) |
Keywords | Field | DocType |
inpainting,face,facial image,Level-of-Detail Reconstruction Model,Restricted Boltzmann Machine | Restricted Boltzmann machine,Computer vision,Pattern recognition,Computer science,Inpainting,Artificial intelligence,Decoding methods,Encoding (memory) | Conference |
Citations | PageRank | References |
0 | 0.34 | 7 |
Authors | ||
6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Christoph Jansen | 1 | 11 | 4.86 |
Radek Mackowiak | 2 | 15 | 1.72 |
Nico Hezel | 3 | 28 | 6.05 |
Moritz Ufer | 4 | 0 | 0.34 |
Gregor Altstadt | 5 | 0 | 0.34 |
Kai Uwe Barthel | 6 | 120 | 17.86 |