2020 BRN Discussion, page-26917

And here is the last one for the day. This research paper appears on Research Gate. The four authors do not give away any details of who they work for or what their respective qualifications are and this appears to be their one and only collaboration. Some investigative work by myself has not revealed anything that I can reliably say about Chuanzhao Han or Jiankun Chen however I am on balance satisfied that Yirong Wu is employed by Honeywell in the USA and Xiaolan Qiu is employed by General Motors in the USA.

The details available to me on Researchgate are limited but I enclose everything that I can access below. To get more information you need to subscribe and be approved for it to be released to you:

Unsupervised Learning Method for SAR Image Classification Based onSpiking NeuUral Network

• January 2021

DOI: 10.20944/preprints202102.0083.v1

Authors:

SHAPE \* MERGEFORMAT

Jiankun Chen

SHAPE \* MERGEFORMAT

Xiaolan Qiu

SHAPE \* MERGEFORMAT

Chuanzhao Han

SHAPE \* MERGEFORMAT

Yirong Wu

Request file PDF

To read the file of this research, you can request a copy directly fromthe authors.

Preprints andearly-stage research may not have been peer reviewed yet.

Download citation

Copy link

References (3)

Abstract

Recent neuroscienceresearch results show that the nerve information in the brain is not onlyencoded by the spatial information. Spiking neural network based on pulsefrequency coding plays a very important role in dealing with the problem ofbrain signal, especially complicated space-time information. In this paper, anunsupervised learning algorithm for bilayer feedforward spiking neural networksbased on spike-timing dependent plasticity (STDP) competitiveness is proposedand applied to SAR image classification on MSTAR for the first time. The SNNlearns autonomously from the input value without any labeled signal and theoverall classification accuracy of SAR targets reached 80.8%. The experimentalresults show that the algorithm adopts the synaptic neurons and networkstructure with stronger biological rationality, and has the ability to classifytargets on SAR image. Meanwhile, the feature map extraction ability of neuronsis visualized by the generative property of SNN, which is a beneficial attemptto apply the brain-like neural network into SAR image interpretation.