...
首页> 外文期刊>電子情報通信学会技術研究報告. ニュ-ロコンピュ-ティング. Neurocomputing >Real-time individual discrimination system by association cortex-entorhinal cortex-Hippocampal formation model
【24h】

Real-time individual discrimination system by association cortex-entorhinal cortex-Hippocampal formation model

机译:皮层-内层皮层-海马形成模型实时个体识别系统

获取原文
获取原文并翻译 | 示例
           

摘要

In the brain, it is thought that the shape recognition system and the spatial recognition system operates in parallel. An association cortex - entorhinal cortex - hippocampal formation model (AEH model) is constructed based on the neurophysiological knowledge and Jones's hypothesis. In the model, the inhibitory action from the inferotemporal cortex to parietal cortex has a gating rule for the activation of the entorhinal cortex and hippocampal formation. Using the AEH model, we developed a real-time individual discrimination system which operated similar to human perception. The whole face was detected from the input picture of a digital video-camera using a real-time eye tracking system. The separation of the whole face into four face parts (eyebrows, eyes, nose, mouth) was automated. We presented the separated face parts picture to the AEH model. We regarded that the AEH model recognized each face correctly if the AEH model learned the face the first time and recollected it the second time. In the experiment, the recognition performance of the model was 100% correct. Besides, long-term potentiation of the memory occured in recollection. In results, the AEH model learned new faces additionally without destroying old faces memory, even if many new faces were presented successively.
机译:在大脑中,认为形状识别系统和空间识别系统并行地工作。基于神经生理学知识和琼斯假说,建立了皮层-内嗅皮层-海马形成模型(AEH模型)。在该模型中,从颞下皮质到顶叶皮质的抑制作用具有激活内嗅皮质和海马形成的门控规则。使用AEH模型,我们开发了一种实时的个体歧视系统,其运行方式类似于人类的感知。使用实时眼动跟踪系统从数字摄像机的输入图片中检测出整个脸部。自动将整个脸分为四个脸部(眉毛,眼睛,鼻子,嘴巴)。我们将分离的面部部分图片呈现给AEH模型。我们认为,如果AEH模型第一次学习到该面孔并第二次对其进行收集,则AEH模型可以正确识别每个面孔。在实验中,模型的识别性能是100%正确的。此外,记忆的长期增强发生在回忆中。结果,即使连续出现了许多新面孔,AEH模型也可以在不破坏旧面孔记忆的情况下额外学习新面孔。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号