What's in a Face? Metric Learning for Face Characterization
| dc.contributor.author | Sendik, Omry | en_US |
| dc.contributor.author | Lischinski, Dani | en_US |
| dc.contributor.author | Cohen-Or, Daniel | en_US |
| dc.contributor.editor | Alliez, Pierre and Pellacini, Fabio | en_US |
| dc.date.accessioned | 2019-05-05T17:41:52Z | |
| dc.date.available | 2019-05-05T17:41:52Z | |
| dc.date.issued | 2019 | |
| dc.description.abstract | We present a method for determining which facial parts (mouth, nose, etc.) best characterize an individual, given a set of that individual's portraits. We introduce a novel distinctiveness analysis of a set of portraits, which leverages the deep features extracted by a pre-trained face recognition CNN and a hair segmentation FCN, in the context of a weakly supervised metric learning scheme. Our analysis enables the generation of a polarized class activation map (PCAM) for an individual's portrait via a transformation that localizes and amplifies the discriminative regions of the deep feature maps extracted by the aforementioned networks. A user study that we conducted shows that there is a surprisingly good agreement between the face parts that users indicate as characteristic and the face parts automatically selected by our method. We demonstrate a few applications of our method, including determining the most and the least representative portraits among a set of portraits of an individual, and the creation of facial hybrids: portraits that combine the characteristic recognizable facial features of two individuals. Our face characterization analysis is also effective for ranking portraits in order to find an individual's look-alikes (Doppelgängers). | en_US |
| dc.description.number | 2 | |
| dc.description.sectionheaders | Learning Images | |
| dc.description.seriesinformation | Computer Graphics Forum | |
| dc.description.volume | 38 | |
| dc.identifier.doi | 10.1111/cgf.13647 | |
| dc.identifier.issn | 1467-8659 | |
| dc.identifier.pages | 405-416 | |
| dc.identifier.uri | https://doi.org/10.1111/cgf.13647 | |
| dc.identifier.uri | https://diglib.eg.org:443/handle/10.1111/cgf13647 | |
| dc.publisher | The Eurographics Association and John Wiley & Sons Ltd. | en_US |
| dc.subject | facial hybrids | |
| dc.subject | face recognition | |
| dc.subject | feature polarization | |
| dc.subject | neural networks CCS Concepts | |
| dc.subject | Computing methodologies | |
| dc.subject | Neural networks | |
| dc.subject | Image processing | |
| dc.title | What's in a Face? Metric Learning for Face Characterization | en_US |