Analog VLSI Neural Circuits CS599 – computational architectures in biological vision

  • View
    222

  • Download
    0

Embed Size (px)

Text of Analog VLSI Neural Circuits CS599 – computational architectures in biological vision

  • Slide 1
  • Analog VLSI Neural Circuits CS599 computational architectures in biological vision
  • Slide 2
  • Charge-Coupled Devices Uniform array of sensors Very little on-board processing Very inexpensive
  • Slide 3
  • CMOS devices More onboard processing Even cheaper! Example: ICM532B from www.ic- media.com: single-chip solution includes photoreceptor array, various gain control and color adjustment mechanisms, image compression and USB interface. Just add a lens and provide power!www.ic- media.com
  • Slide 4
  • The challenge Digital processing is power hungry Analog processing is much more energy efficient But so much variability in the gain of transistors obtained when fabricating highly integrated (VLSI) chips that analog computations seem impossible: nearly each analog amplifier on the chip should be associated with control pins, analog memories, etc to correct for fabrication variability. Hopeless situation?
  • Slide 5
  • A VLSI MOS transistor
  • Slide 6
  • An analog chip layout: the wish
  • Slide 7
  • An actual chip: the cold reality
  • Slide 8
  • Biological motivation Well, there is also a lot of variability in size and shape of neurons from a same class But the brain still manages to produce somewhat accurate computations Whats the trick? online adaptability to counteract morphological and electrical mismatches among elementary components.
  • Slide 9
  • Remember? Electron Micrograph of a Real Neuron
  • Slide 10
  • Mahowald & Meads Silicon Retina Smoothing network: allows system to adapt to various light levels.
  • Slide 11
  • Andreou and Boahen's silicon retina See http://www.iee.et.tu-dresden.de/iee/eb/ analog/papers/mirror/visionchips/vision_chips/ andreou_retina.html
  • Slide 12
  • Diffusive network dQn/dt is the current supplied by the network to node n, and D is the diffusion constant of the network, which depends on the transistor parameters, and the voltage Vc.
  • Slide 13
  • Full network Two layers of the diffusive network: upper corresponds to horizontal cells in retina and lower to cones. Horizontal N- channel transistors model chemical synapses. The function of the network can be approximated by the biharmonic equation where g and h are proportional to the diffusivity of the upper and lower smoothing layers, respectively.
  • Slide 14
  • Full network
  • Slide 15
  • VLSI sensor with retinal organization
  • Slide 16
  • Carver Mead: the floating gate www.cs.washington.edu/homes/hsud/fg_workshop.html
  • Slide 17
  • Spatial layout
  • Slide 18
  • Electron tunneling
  • Slide 19
  • Slide 20
  • Hot electron injection
  • Slide 21
  • Slide 22
  • Spatial layout
  • Slide 23
  • A learning synapse circuit

Recommended

View more >