It is used to take image prior into account to fastly and accurately predict the image details and to make the neurons work on the local image patches to avoid compromise to different image contents.
On the other hand, in all the previous deep-learning-based methods, each neuron works on the whole input feature map.
That results in each neuron works on the same kinds of image patches to avoid compromise to different image contents.
Neuron works with fixed point arithmetic, and is designed with the accent to effective utilization of the FPGA resources, parallelism and computing speed.
In the brain, these neurons work together in complex networks, and so understanding how the brain processes information will require neuroscientists to analyze voltage changes in many neurons at the same time.To achieve this, scientists have developed genetically-encoded voltage indicators (or GEVIs).
How a biological neuron works is modeled by an artificial neuron, as depicted by Figure1.
"Richard pulled a sea slug, Aplysia californica, out of a salt water tank and suggested to me that it would be a good idea to start by understanding how a single neuron works," Kreiner says.
In the wake of OSTP's July 2012 request for proposals for "Grand Challenges" in science and technology similar to the Human Genome Project, Chun and other co-authors of the Neuron paper worked with officials at federal agencies particularly OSTP Deputy Director for Policy Tom Kalil to refine the proposal, tailoring it to focus more on human applications.
He chose to work on the principal cell type of the dopamine-making region of the brain, known as the medium-spiny projection neuron, because unlike the cerebral cortex, where many different types of neuron work side by side, the corpus striatum has almost pure medium-spiny neurons, giving enough material for biochemical analysis.
