** 最新成果:He H, Yang X, Xu Z, Deng N, Shang Y, et al. (2019) Implementing artificial neural networks through bionic construction. PLOS ONE 14(2): e0212368. https://doi.org/10.1371/journal.pone.0212368;Xu Yang, Guo Liu, Songgaojun Deng, Zichao Wei, Hu He*, Yingjie Shang, Ning Deng. Artif Intell Rev (2018). https://doi.org/10.1007/s10462-018-9626-2
** 最新成果:He H, Yang X, Xu Z, Deng N, Shang Y, et al. (2019) Implementing artificial neural networks through bionic construction. PLOS ONE 14(2): e0212368. https://doi.org/10.1371/journal.pone.0212368; Xu Yang, Guo Liu, Songgaojun Deng, Zichao Wei, Hu He*, Yingjie Shang, Ning Deng. Artif Intell Rev (2018). https://doi.org/10.1007/s10462-018-9626-2
实现了模拟电路隐蔽型硬件木马和抗硬件木马设计(R2D2: Runtime Reassurance and Detection of A2 Trojan Accepted by HOST 2018 and TCAD)
集成了嵌入式1Mb的ReRAM
Egret(RISC-V CPU) coming soon.
Evolution Neural Network
仿生型自生长神经网
依靠计算能力实现的人工神经网不具备强(通用)人工智能的能力
生物神经网既有精巧的信号处理网络,也有鲁棒性高,通用性好的学习网络
生物网络天生具备因果学习机制,能够进行因果的学习和推理
生物神经网络具备后天生长特性,可以增加神经网络原来不具备的功能
生物神经网的记忆是分布式的,采用内容提取记忆的方式
最新成果:He H, Yang X, Xu Z, Deng N, Shang Y, et al. (2019) Implementing artificial neural networks through bionic construction. PLOS ONE 14(2): e0212368. https://doi.org/10.1371/journal.pone.0212368; Xu Yang, Guo Liu, Songgaojun Deng, Zichao Wei, Hu He*, Yingjie Shang, Ning Deng. Artif Intell Rev (2018). https://doi.org/10.1007/s10462-018-9626-2
论文:CDSP: A Solution for Privacy and Security of Multimedia Information Processing in Industrial Big Data and Internet of Things, https://doi.org/10.3390/s19030556