Multi-view perceptron: A deep model for learning face identity and view representations
Refereed conference paper presented and published in conference proceedings


Full Text

Times Cited

Other information
AbstractVarious factors, such as identity, view, and illumination, are coupled in face images. Disentangling the identity and view representations is a major challenge in face recognition. Existing face recognition systems either use handcrafted features or learn features discriminatively to improve recognition accuracy. This is different from the behavior of primate brain. Recent studies [5, 19] discovered that primate brain has a face-processing network, where view and identity are processed by different neurons. Taking into account this instinct, this paper proposes a novel deep neural net, named multi-view perceptron (MVP), which can untangle the identity and view features, and in the meanwhile infer a full spectrum of multi-view images, given a single 2D face image. The identity features of MVP achieve superior performance on the MultiPIE dataset. MVP is also capable to interpolate and predict images under viewpoints that are unobserved in the training data.
All Author(s) ListZhu Z., Luo P., Wang X., Tang X.
Name of Conference28th Annual Conference on Neural Information Processing Systems 2014, NIPS 2014
Start Date of Conference08/12/2014
End Date of Conference13/12/2014
Place of ConferenceMontreal
Country/Region of ConferenceCanada
Detailed descriptionorganized by Neural Information Processing Systems Foundation,
Year2014
Month1
Day1
Volume Number1
Issue NumberJanuary
Pages217 - 225
ISSN1049-5258
LanguagesEnglish-United Kingdom

Last updated on 2021-28-11 at 23:35