Shoko Imaizumi

Bio: Shoko Imaizumi is an academic researcher. The author has an hindex of 1, co-authored 1 publications receiving 36 citations.

国際会議開催報告:IEEE International Symposium on Circuits and Systems

Abstract: ISCASはIEEE Circuits and Systems Societyが主催する 回路とシステムに関する国際会議で,今年は5月20日から5 月23日にかけて,韓国のソウルで開催された. ISCASは,「回路とシステム」に関する国際会議としては最 大規模かつ最も権威のある学会である.近年は,専門性を重視 した国際会議が増える中,ISCASは回路とシステムに関する 総合大会的な会議の立ち位置を維持し続けている.今年の会議 テーマは,“Convergence of BINET (Bio, Info, Nano, and Enviro Technology)”ということで,近年注目されている生物 学やナノ及び環境テクノロジーに関する論文をはじめ,これま で同様にアナログ信号処理,ディジタル信号処理,通信用回路, 非線形回路,パワーエレクトロニクス,VLSIシステムなど多 岐にわたる分野から論文投稿が行われた.テクニカルプログラ ムとしては,14のトラックからプログラムが構成され,102 のレクチャーセッション,52のポスターセッション,及び四 つのライブデモセッションが組まれた.レギュラートラックに 限れば,投稿総数1,651件に対して741件の論文が採録され, 採択率は44.9 %であった.なお,スペシャルセッション,デ モセッションを含めると1,760件の投稿数で822件が採録さ れ,採択率は46.7 %であった.ISCASが目標としている採 択率50 %以下を今年も維持する結果となった.また,3人の 世界的に著名な研究者による基調講演が会議のテクニカルプロ グラムの質を更に向上させることになった. ソーシャルイベントもかなり充実していた.初日のウェルカ ムレセプションでは,氷でできたISCAS2012の彫刻が我々 参加者を迎えてくれ,世界各国の料理が振る舞われた. また,会場ステージでは,CASSメンバーによって結成され たCASSバンドの演奏があり,大いに盛り上がりを見せた.三 日目のバンケットでは,韓国の伝統芸能と,K-POPスターの 4MINUTEのステージがあり,韓国のエンターテイメントのす ごさを実感することができた.最終日のフェアウェルパーティ は,韓国国立博物館で開催され,まず博物館で韓国の歴史や 文化を学んだあと,参加者との最後の別れを交わすパーティが 始まった.ここでは,巨大ビビンバやマッコリが振る舞われ, 韓国の食文化を存分に味わうことができた.このようにして, ISCAS 2012はテクニカルプログラム及びソーシャルプログ ラム共に大成功して幕を閉じた. 来年のISCAS 2013は5月19日~23日に中国の北京で 開催されます.私はもちろん参加します!!皆さん,来年中国 でお会いしましょう!!

Nanotechnology-Enabled Energy Harvesting for Self-Powered Micro-/Nanosystems

Abstract: Health, infrastructure, and environmental monitoring as well as networking and defense technologies are only some of the potential areas of application of micro-/nanosystems (MNSs). It is highly desirable that these MNSs operate without an external electricity source and instead draw the energy they require from the environment in which they are used. This Review covers various approaches for energy harvesting to meet the future demand for self-powered MNSs.

Energy harvesting from low frequency applications using piezoelectric materials

Abstract: In an effort to eliminate the replacement of the batteries of electronic devices that are difficult or impractical to service once deployed, harvesting energy from mechanical vibrations or impacts using piezoelectric materials has been researched over the last several decades. However, a majority of these applications have very low input frequencies. This presents a challenge for the researchers to optimize the energy output of piezoelectric energy harvesters, due to the relatively high elastic moduli of piezoelectric materials used to date. This paper reviews the current state of research on piezoelectric energy harvesting devices for low frequency (0–100 Hz) applications and the methods that have been developed to improve the power outputs of the piezoelectric energy harvesters. Various key aspects that contribute to the overall performance of a piezoelectric energy harvester are discussed, including geometries of the piezoelectric element, types of piezoelectric material used, techniques employed to match the resonance frequency of the piezoelectric element to input frequency of the host structure, and electronic circuits specifically designed for energy harvesters.

Synchronization dynamics on the picosecond time scale in coupled Josephson junction neurons

Abstract: Conventional digital computation is rapidly approaching physical limits for speed and energy dissipation. Here we fabricate and test a simple neuromorphic circuit that models neuronal somas, axons, and synapses with superconducting Josephson junctions. The circuit models two mutually coupled excitatory neurons. In some regions of parameter space the neurons are desynchronized. In others, the Josephson neurons synchronize in one of two states, in-phase or antiphase. An experimental alteration of the delay and strength of the connecting synapses can toggle the system back and forth in a phase-flip bifurcation. Firing synchronization states are calculated g70 000 times faster than conventional digital approaches. With their speed and low energy dissipation (${10}^{\ensuremath{-}17}\phantom{\rule{0.16em}{0ex}}\mathrm{J}/\mathrm{spike}$), this set of proof-of-concept experiments establishes Josephson junction neurons as a viable approach for improvements in neuronal computation as well as applications in neuromorphic computing.

Pulse-Doppler Signal Processing with Quadrature Compressive Sampling

Abstract: Quadrature compressive sampling (QuadCS) is a newly introduced sub-Nyquist sampling for acquiring inphase and quadrature (I/Q) components of radio-frequency signals. For applications to pulse-Doppler radars, the QuadCS outputs can be arranged in 2-dimensional data similar to that by Nyquist sampling. This paper develops a compressive sampling pulse-Doppler (CoSaPD) processing scheme from the sub-Nyquist samples. The CoSaPD scheme follows Doppler estimation/detection and range estimation and is conducted on the sub-Nyquist samples without recovering the Nyquist samples. The Doppler estimation is realized through spectrum analyzer as in classic processing. The detection is done on the Doppler bin data. The range estimation is performed through sparse recovery algorithms on the detected targets and thus the computational load is reduced. The detection threshold can be set at a low value for improving detection probability and then the introduced false targets are removed in the range estimation stage through inherent detection characteristic in the recovery algorithms. Simulation results confirm our findings. The CoSaPD scheme with the data at one eighth the Nyquist rate and for SNR above -25dB can achieve performance of the classic processing with Nyquist samples.

Pruning of Deep Spiking Neural Networks through Gradient Rewiring.

Abstract: Spiking Neural Networks (SNNs) have been attached great importance due to their biological plausibility and high energy-efficiency on neuromorphic chips. As these chips are usually resource-constrained, the compression of SNNs is thus crucial along the road of practical use of SNNs. Most existing methods directly apply pruning approaches in artificial neural networks (ANNs) to SNNs, which ignore the difference between ANNs and SNNs, thus limiting the performance of the pruned SNNs. Besides, these methods are only suitable for shallow SNNs. In this paper, inspired by synaptogenesis and synapse elimination in the neural system, we propose gradient rewiring (Grad R), a joint learning algorithm of connectivity and weight for SNNs, that enables us to seamlessly optimize network structure without retraining. Our key innovation is to redefine the gradient to a new synaptic parameter, allowing better exploration of network structures by taking full advantage of the competition between pruning and regrowth of connections. The experimental results show that the proposed method achieves minimal loss of SNNs' performance on MNIST and CIFAR-10 dataset so far. Moreover, it reaches a $\sim$3.5% accuracy loss under unprecedented 0.73% connectivity, which reveals remarkable structure refining capability in SNNs. Our work suggests that there exists extremely high redundancy in deep SNNs. Our codes are available at this https URL.