Neurons, N-tuples and faces

Buy article PDF
© The Institution of Electrical Engineers
Published 21/07/2012

Abstract

Enthusiasm for neural computing seems, at present, to know no bounds. Neural networks are powerful pattern classifiers that are normally thought to require custom VLSI devices in order to obtain the advantages of their inherent parallelism. By exploiting N-tuple techniques, the real-time application of such networks is possible, using conventional computers, for such tasks as recognising and locating facial features.

Inspec keywords: neurophysiology; parallel processing; neural nets; pattern recognition

Other keywords: conventional computers; facial features; custom VLSI devices; pattern classifiers; N-tuple techniques; real-time application; neurons; inherent parallelism; neural computing

Subjects: Artificial intelligence; Systems theory applications in biology and medicine; Multiprocessing systems; Pattern recognition

References

    1. 1)
    2. 2)
      • Bledsoe, W.W., Browning, I.: Proc. Eastern Joint Computer Conference, 1959, p. 225–232
    3. 3)
      • Kanerva, P.: , Sparse distributed memory, 1988 (MIT PressCambridge, USA)
    4. 4)
      • Aleksander, I.: , Advanced digital information systems, 1985 (Prentice-Hall InternationalLondon)
    5. 5)
      • Johnson, M.J., Allinson, N.M., Moon, K.J., Touretzky, D.S.: , Advances in neural information processing systems, 1989 (Morgan KaufmannSan Mateo)
    6. 6)
      • Allinson, N.M., Johnson, M.J., Taylor, J.G., Mannion, C.L.T.: , New developments in neural computing, 1989 (Adam HilgerBristol)
    7. 7)
      • Luckman, A.J., Allinson, N.M.: Proc. SPIE, 1990, 1197
Print this page
This is a required field
Please enter a valid email address