卒業生とその進路

3 次元ニューロモルフィックネットワークのためのランダムウェイト消去を用いたスマートアーキテクチャ


アリ エミリアーノ ホセ

2021 年度 卒 /修士(情報科学)

修士論文の概要

Artificial Neural Networks (ANNs) are being used in many areas of applications in the information science field with the software version of them being the most widely used and, as a consequence, the most widely studied and optimized. Due to this, hardware ANNs still have a lot of room to grow and change in order to achieve a widespread use since dealing with physical devices albeit have some advantages regarding the processing speed, the cost and space limitations are the main issue when compared to their software counterparts. Different devices are being used for the hardware implementation of networks and the use of electropolymerization of conductive wires is one of the promising ones. These devices, called Molecular Synapses onward, have two advantages that make them attractive. First, no pre-fabrication is needed apart from passive electrodes, the wire is then grown between two electrodes in order to create a conductive path. This means that if the network were to use a low number of connections, no waste is generated since only the needed material would be grown. Second, the devices are not limited to 1 dimensional space, currently 2 dimensional devices have been used for the implementation of working neural networks and the possibility of working 3 dimensional neuromorphic devices exists. Although the 3D connectivity is interesting in order to mimic our brain structure, this is no easy task when an additional problem is presented: in hardware ANNs is the duplication of connections due to the impossibility of creating negative conductances, and thus the need of differential structures to represent each weight. This results in denser connections and a higher chance of wires crossing. In this thesis, 2D experiments of Molecular Synapses are shown and explained in order to set the basis of the current working devices and the problems that they have. From that basis several 3D implementation proposals will be showcased with their possible issues associated to the electrode fabrication and wire growth. Once that the 3D devices have been explained as a possibility, then a smart architecture using random weight elimination for the reduction of the number of the connections in a network will be explained for the general case and then applied to a 3D simulation case to show the possible density reduction.

学術論文

  1. Amemiya Y., Ali E.J., Hagiwara N., Akai-Kasaya M., and Asai T., "Heuristic model for configurable polymer wire synaptic devices," Nonlinear Theory and Its Applications, vol. E13-N, no. 2, pp. 379-384 (2022).
  2. Ali E.J., Amemiya Y., Akai-Kasaya M., and Asai T., "Smart hardware architecture with random weight elimination and weight balancing algorithms," Nonlinear Theory and Its Applications, vol. E13-N, no. 2, pp. 336-342 (2022).

招待講演/セミナー

  1. Ali E.J., Amemiya Y., Hagiwara N., Akai-Kasaya M., and Asai T., "A comparison between simulations and experiments of neuromorphic devices using electropolymerization of conductive polymer nanowires," Joint Symposium of JSPS-DST Bilateral Research on Charge- and Spin-Blockade in Ultrathin-Layers of Single Molecule Magnets, online, Japan (Feb. 24, 2021).

国際会議

  1. Hagiwara N., Amemiya Y., Ali E.J., Asai T., and Akai-Kasaya M., "Feasibility of neuromorphic wetware using configurable polymer networks," The 10th RIEC International Symposium on Brain Functions and Brain Computer, Online, (Feb. 18-19, 2022).
  2. Amemiya Y., Ali E.J., Hagiwara N., Akai-Kasaya M., and Asai T., "A heuristic model for configurable polymer-wire synaptic devices," The 2021 Nonlinear Science Workshop, Online (Dec. 6-8, 2021).
  3. Ali E.J., Amemiya Y., Akai-Kasaya M., and Asai T., "Smart hardware architecture with random weight elimination and weight balancing algorithms," The 2021 Nonlinear Science Workshop, Online (Dec. 6-8, 2021).
  4. Ali E.J., Asai T., and Akai-Kasaya M., "Trainer system for the study of hardware artificial neural networks," RISP International Workshop on Nonlinear Circuits, Communications and Signal Processing 2021, Online (Mar. 1-3, 2021).
  5. Amemiya Y., Ali E.J., Akai-Kasaya M., and Asai T., "A cellular automata model for unstructured, flexible, and configurable molecular synapses toward edge-AI computing," 2020 International Symposium on Nonlinear Theory and Its Applications, Online, Japan (Nov. 16-19, 2020).

国内学会

  1. 萩原 成基, 雨宮 佳希, アリ ホセ エミリアーノ, 浅井 哲也, 赤井 恵, "立体配線型メモリ素子で構成される新規脳型回路アーキテクチャの検討," 電子情報通信学会複雑コミュニケーションサイエンス研究会, 北海道 ルスツリゾートホテル&コンベンション, (ハイブリッド開催), 2022年3月27日.
  2. 雨宮 佳希, アリ ホセ エミリアーノ, 赤井 恵, 浅井 哲也, "再構成可能な分子シナプス素子の簡易モデル," 第34回 回路とシステムワークショップ, 北九州国際会議場, (小倉), 2021年8月26-27日.
  3. アリ ホセ エミリアーノ, 雨宮 佳希, 浅井 哲也, 赤井 恵, "Neuromorphic Devices using Spatial Free Wiring of Conductive Polymer for Hardware Artificial Neural Networks," 電子情報通信学会複雑コミュニケーションサイエンス研究会, (オンライン開催), 2021年3月29日.