Results 111 to 120 of about 79,032 (283)

Refined Gate: A Simple and Effective Gating Mechanism for Recurrent Units

, 2020
Recurrent neural network (RNN) has been widely studied in sequence learning tasks, while the mainstream models (e.g., LSTM and GRU) rely on the gating mechanism (in control of how information flows between hidden states). However, the vanilla gates in RNN (e.g., the input gate in LSTM) suffer from the problem of gate undertraining, which can be caused ...
Cheng, Zhanzhan, Xu, Yunlu, Cheng, Mingjian, Qiao, Yu, Pu, Shiliang, Niu, Yi, Wu, Fei   +6 more
openaire   +2 more sources

Designing Asymmetric Memristive Behavior in Proton Mixed Conductors for Neuromorphic Applications

Advanced Functional Materials, EarlyView.
Protonic devices that couple ionic and electronic transport are demonstrated as bioinspired neuromorphic elements. The devices exhibit rubber‐like asymmetric memristive behavior with slow voltage‐driven conductance increase and rapid relaxation, enabling simplified read–write operation.
Nada H. A. Besisa, Masaki Tsuji, Yoshihiro Furue, Yifan Geng, Satya P. Pati, Satoshi Hamasuna, Takeaki Yajima   +6 more
wiley   +1 more source

Trap‐Assisted Transport and Neuromorphic Plasticity in Lead‐Free 2D Perovskites PEA2SnI4

Advanced Functional Materials, EarlyView.
An artificial retina built from lead‐free layered perovskite (PEA)2SnI4 converts light input into a persistent photocurrent and sums successive flashes over time. Micro/nanocrystals integrated on electrodes act as synapse‐like pixels that perform temporal integration directly in hardware. This in‐sensor preprocessing merges detection and computation on
Ofelia Durante, Valeria Demontis, Karl Jonas Riisnaes, Lam Hoi Tung, Sebastiano De Stefano, Daniela Marongiu, Adolfo Mazzotti, Aditya Bhardwaj, Francesco Mattana, Elisa Pili, Selene Matta, Federico Paolucci, Michele Saba, Andrea Mura, Giovanni Bongiovanni, Monica F Craciun, Saverio Russo, Antonio Di Bartolomeo   +17 more
wiley   +1 more source

Mapping Nanoscale Buckling in Atomically Thin Cr2Ge2Te6

Advanced Functional Materials, EarlyView.
Atomic‐resolution STEM is used to resolve nanoscale buckling in monolayer Cr2Ge2Te₆. A noise‐robust image analysis reconstructs three‐dimensional lattice distortions from single plan‐view images, revealing pronounced defect‐driven nm‐scale out‐of‐plane buckling.
Amy Carl, Nicholas Clark, David G. Hopkinson, Matthew Hamer, Matthew Watson, Laxman Nagireddy, James E. Nunn, Alexei Barinov, Yichao Zou, William Thornley, Casey Cheung, Wendong Wang, Sam Sullivan‐Allsop, Xiao Li, Astrid Weston, Eli G. Castanon, Andrey V Kretinin, Cephise Cacho, Neil R. Wilson, Sarah J. Haigh, Roman Gorbachev   +20 more
wiley   +1 more source

Online Persian/Arabic Writer Identification using Gated Recurrent Unit Neural Networks

Majlesi Journal of Electrical Engineering
Conventional methods in writer identification mostly rely on hand-crafted features to represent the characteristics of different handwritten scripts.
Mahsa Aliakbarzadeh, Farbod Razzazi
doaj   +1 more source

Minion Gated Recurrent Unit for Continual Learning

The increasing demand for continual learning in sequential data processing has led to progressively complex training methodologies and larger recurrent network architectures. Consequently, this has widened the knowledge gap between continual learning with recurrent neural networks (RNNs) and their ability to operate on devices with limited memory and ...
Zyarah, Abdullah M., Kudithipudi, Dhireesha   +1 more
openaire   +2 more sources

VIVID: A qPCR‐Based Platform for Sensitive and Quantitative In Vivo Tracking of Extracellular Vesicles

Advanced Functional Materials, EarlyView.
This study introduces VIVID (Vesicle In Vivo Identification using DNA), a qPCR‐based platform that tracks PCR‐amplifiable DNA tags loaded in the EVs for accurate and quantifiable EV biodistribution in vivo. ABSTRACT Extracellular vesicles (EVs) represent promising carriers for nucleic acid therapeutics, offering advantages over synthetic nanoparticles ...
Oscar Boyadjian, Talin Ebrahimian, Jonathan Miranda O'Connor, Janusz Rak, Stephanie Lehoux, Maryam Tabrizian   +5 more
wiley   +1 more source

Non-normal Recurrent Neural Network (nnRNN): learning long time dependencies while improving expressivity with transient dynamics

, 2019
A recent strategy to circumvent the exploding and vanishing gradient problem in RNNs, and to allow the stable propagation of signals over long time scales, is to constrain recurrent connectivity matrices to be orthogonal or unitary.
Bengio, Yoshua, Gidel, Gauthier, Goyette, Kyle, Kerg, Giancarlo, Lajoie, Guillaume, Touzel, Maximilian Puelma, Vorontsov, Eugene   +6 more
core  

Bio‐Inspired Molecular Events in Poly(Ionic Liquids)

Advanced Functional Materials, EarlyView.
Originating from dipolar and polar inter‐ and intra‐chain interactions of the building blocks, the topologies and morphologies of poly(ionic liquids) (PIL) govern their nano‐ and micro‐processibility. Modulating the interactions of cation‐anion pairs with aliphatic dipolar components enables the tunability of properties, facilitated by “bottom‐up ...
Jiahui Liu, Marek W. Urban
wiley   +1 more source

Unraveling Quantitative Sensing Mechanism and Predictive Molecular Metrics for High‐Performance OFET Amine Sensors

Advanced Functional Materials, EarlyView.
This study introduces a novel chloro boron subphthalocyanine/polymer blend OFET sensor achieving 0.005 ppb limit of detection for ammonia at room temperature and high selectivity against similar amines. An original theoretical framework is proposed to describe the sensing mechanism, relating analyte molecular volume and Lewis basicity to sensor ...
Kavinraaj Ella Elangovan, Matthew P. Confer, Rui Zhang, Siqing Wang, Jeffrey S. Moore, Rohit Bhargava, Ying Diao   +6 more
wiley   +1 more source