Cool Camera Tech: Retina-Like Sensors and Brain-Like Computers

Posted on February 17, 2015

0


Technology companies keep pumping out the high tech wonders and two technologies, from IBM, have the ability to make you sit up and say ‘Wow’.

IBM has developed a camera called the DVS128 in collaboration with Cornell Tech and research laboratory iniLabs that operates like the human eye. Taking their cues from biology, the iniLabs team recognised that the eye requires far less power than a digital camera and is more efficient at processing images. Its digital visual sensor (DVS) system does away with the idea of seeing the world as a series of frames.

“Successive frames contain enormously redundant information, wasting energy, computational power and time. In addition, each frame imposes the same exposure time on every pixel, making it impossible to process scenes containing very dark and very bright regions,” says iniLabs. By operating like the human retina the DVS reduces power, data storage and computational requirements drastically, while increasing dynamic sensor range by orders of magnitude. No images are sent out of entire images at fixed frame rates. “Only the local pixel-level changes caused by moving in a scene are transmitted – at exactly the time they occur. The result is a stream of events at microsecond time resolution, equivalent to or better than conventional high-speed vision sensors running at thousands of frames per second.”

iniLabs says the camera could be used for:

  • Surveillance and ambient sensing
  • Fast robotics for mobile (eg, drone video) and fixed sources (eg the RoboGoalie video below)
  • Factory automation
  • Hydrodynamics
  • Microscopy
  • Motion analysis (eg, human or animal motion)
  • Sleep research and chronobiology
  • Fluorescent imaging
  • Particle Tracking

 

Hitting TrueNorth
DVS works with IBM’s brain-like computing architecture called TrueNorth, which is described in the MIT Technology Review as an approach that “stores and processes information in a distributed, parallel way, like the neurons and synapses in a brain.” IBM’s partnership with iniLabs and Cornell has also pioneered the first ‘neurosynaptic’ chip, with one million programmable neurons and 256 million programmable synapses, capable of 46 billion synaptic operations per second.

IBM synaptic chip
The postage stamp-sized chip is much more energy efficient than regular chips, only consuming the same amount of power as a typical hearing aid. IBM says a synaptic chip in a mobile device could make any smartphone as powerful as a supercomputer. Programming the chips requires specialist knowledge, leading IBM to develop its own curriculum for students called the Synapse University.

Advertisements
Posted in: Camera, Video