Results 51 to 60 of about 8,125 (267)

Lightweight Deep Neural Network Embedded with Stochastic Variational Inference Loss Function for Fast Detection of Human Postures

Entropy, 2023
Fusing object detection techniques and stochastic variational inference, we proposed a new scheme for lightweight neural network models, which could simultaneously reduce model sizes and raise the inference speed.
Feng-Shuo Hsu, Zi-Jun Su, Yamin Kao, Sen-Wei Tsai, Ying-Chao Lin, Po-Hsun Tu, Cihun-Siyong Alex Gong, Chien-Chang Chen   +7 more
doaj   +1 more source

A New Stochastic Inner Product Core Design for Digital FIR Filters

MATEC Web of Conferences, 2017
Stochastic computing (SC) is a computational technique with computational operations governed by probability instead of arithmetic rules. It recently found promising applications in digital and image processing areas and attracted attentions of ...
Wong Ming Ming, Wong M. L. Dennis, Zhang Cishen, Hijazin Ismat   +3 more
doaj   +1 more source

A Stochastic Computational Approach for the Analysis of Fuzzy Systems

IEEE Access, 2017
Fault tree analysis (FTA) has been widely utilized as a reliability evaluation technique for complex systems, such as nuclear power plants and aerospace systems. However, it is hard to obtain the crisp failure probabilities of basic events, owning to the
Xiaogang Song, Zhengjun Zhai, Peican Zhu, Jie Han   +3 more
doaj   +1 more source

An Implementation Method Using Cut-Off Bits for Restricted Boltzmann Machines Without Random Number Generators

IEEE Access, 2022
This study proposes an implementation method of a hardware-oriented restricted Boltzmann machine (RBM) without random number generators (RNGs) that employ cut-off bits, which are obtained from fixed-point binary arithmetic operations on digital hardware,
Sansei Hori, Hakaru Tamukoh
doaj   +1 more source

All‐in‐One Analog AI Hardware: On‐Chip Training and Inference with Conductive‐Metal‐Oxide/HfOx ReRAM Devices

Advanced Functional Materials, EarlyView.
An all‐in‐one analog AI accelerator is presented, enabling on‐chip training, weight retention, and long‐term inference acceleration. It leverages a BEOL‐integrated CMO/HfOx ReRAM array with low‐voltage operation (<1.5 V), multi‐bit capability over 32 states, low programming noise (10 nS), and near‐ideal weight transfer.
Donato Francesco Falcone, Victoria Clerico, Wooseok Choi, Tommaso Stecconi, Folkert Horst, Laura Bégon‐Lours, Matteo Galetta, Antonio La Porta, Nikhil Garg, Fabien Alibart, Bert Jan Offrein, Valeria Bragaglia   +11 more
wiley   +1 more source

A Novel Technique to Control the Accuracy of a Nonlinear Fractional Order Model of COVID-19: Application of the CESTAC Method and the CADNA Library

Mathematics, 2021
In this paper, a nonlinear fractional order model of COVID-19 is approximated. For this aim, at first we apply the Caputo–Fabrizio fractional derivative to model the usual form of the phenomenon.
Samad Noeiaghdam, Sanda Micula, Juan J. Nieto   +2 more
doaj   +1 more source

An Integrative Strategy Delineates Modular Metabolic Remodeling and Potential Therapeutic Targets Across Metabolic Diseases

Advanced Science, EarlyView.
An integrative single‐cell atlas across multiple metabolic diseases reveals coordinated metabolic modules and disease‐shared versus disease‐specific pathway activities. By systematically comparing scoring strategies, a robust RankAve framework is established. Coupled with network analysis and drug‐target prediction, this resource uncovers cross‐disease
Kuan Yang, Congxue Hu, Hang Zhao, Xiao Peng, Yiming Yan, Jiasi Wang, Wenqi Jiang, Depeng Mu, Yan Cao, Xia Li, Yunpeng Zhang   +10 more
wiley   +1 more source

People Counting and Positioning Using Low‐Resolution Infrared Images for FeFET‐Based In‐Memory Computing

Advanced Electronic Materials, EarlyView.
In this work, low‐resolution infrared imaging is combined with a 28 nm FeFET IMC architecture to enable compact, energy‐efficient edge inference. MLC FeFET devices are experimentally characterized, and controlled multi‐level current accumulation is validated at crossbar array level.
Alptekin Vardar, Li Zhang, Alexander Rehmer, Marc Lupp, Subhashree Baskaran, Osama Abdelaal, Franz Müller, Nandakishor Yadav, Bing Li, Thomas Kämpfe   +9 more
wiley   +1 more source

SigmaFormer: Augmenting transformer encoders with COSMO sigma profiles for pure component property prediction

AIChE Journal, EarlyView.
Abstract Transformer‐based molecular models pretrained on SMILES strings demonstrate strong performance in property prediction. However, these model often lack explicit integration of molecular surface charge distributions that govern intermolecular interactions such as hydrogen bonding and polarity.
Tae Hyun Kim, Silabrata Pahari, Joseph Sang‐Il Kwon   +2 more
wiley   +1 more source

Calibration‐Free Electromyography Motor Intent Decoding Using Large‐Scale Supervised Pretraining

Advanced Intelligent Systems, EarlyView.
Calibration‐free electromyography motor intent decoding is enabled through large‐scale supervised pretraining across heterogeneous datasets. A Spatially Aware Feature‐learning Transformer processes variable channel counts and electrode geometries, allowing transfer across users and recording setups. On a held‐out benchmark, fine‐tuned cross‐user models
Alexander E. Olsson, Nebojša Maleševic, Anders Björkman, Christian Antfolk   +3 more
wiley   +1 more source