Subspace-Based Face Recognition in Analog VLSI

Abstract

We describe an analog-VLSI neural network for face recognition based on subspace methods. The system uses a dimensionality-reduction network whose coefficients can be either programmed or learned on-chip to per- form PCA, or programmed to perform LDA. A second network with user- programmed coefficients performs classification with Manhattan distances. The system uses on-chip compensation techniques to reduce the effects of device mismatch. Using the ORL database with 12x12-pixel images, our circuit achieves up to 85% classification performance (98% of an equivalent software implementation).

Cite

Text

Carvajal et al. "Subspace-Based Face Recognition in Analog VLSI." Neural Information Processing Systems, 2007.

Markdown

[Carvajal et al. "Subspace-Based Face Recognition in Analog VLSI." Neural Information Processing Systems, 2007.](https://mlanthology.org/neurips/2007/carvajal2007neurips-subspacebased/)

BibTeX

@inproceedings{carvajal2007neurips-subspacebased,
  title     = {{Subspace-Based Face Recognition in Analog VLSI}},
  author    = {Carvajal, Gonzalo and Valenzuela, Waldo and Figueroa, Miguel},
  booktitle = {Neural Information Processing Systems},
  year      = {2007},
  pages     = {225-232},
  url       = {https://mlanthology.org/neurips/2007/carvajal2007neurips-subspacebased/}
}