Results 181 to 190 of about 61,624 (324)
This work harnesses nonidealities in analog in‐memory computing (IMC) by training physical neural networks modeled with ordinary differential equations. A differentiable spike‐time discretization accelerates training by 20× and reduces memory usage by 100×, enabling large IMC‐equivalent models to learn the CIFAR‐10 dataset.
Yusuke Sakemi +5 more
wiley +1 more source
Neuromorphic Device Based on Material and Device Innovation toward Multimode and Multifunction
In the era of big data, multimodal and multifunctional neuromorphic devices offer significant opportunities in designing AI hardware. In this work, recent advances about emerging material systems and novel device structures in this field are summarized in detail. Potential applications are reviewed.
Feng Guo +3 more
wiley +1 more source
Haptic Perception via the Dynamics of a Flexible Body Inspired by an Ostrich's Neck
Inspired by avian anatomy, this study uses a flexible robotic neck to investigate haptic perception driven by musculoskeletal dynamics. By applying physical reservoir computing, the robot encodes external force interactions into its body dynamics, allowing effective object classification.
Kazashi Nakano +3 more
wiley +1 more source
Quantum medicine: A quantum–mechanical framework for redox biology, disease and precision medicine
Quantum medicine integrates quantum simulation, experimental validation and clinical data into a unified QEC pipeline. Quantum algorithms compute electronic and spin‐level properties of redox biology, which are validated by quantum sensing and structural biology and translated into precision diagnostics and therapy through multi‐omics and quantum ...
Ji‐Yong Sung, Jae‐Ho Cheong
wiley +1 more source
Exploiting Vision Transformer and Ensemble Learning for Advanced Malware Classification
Overview of the proposed RF–ViT ensemble for multi‐class malware classification. Textual (BoW/byte‐frequency) and visual representations are combined via a product rule, achieving improved accuracy and robustness over individual models. ABSTRACT Malware remains a significant concern for modern digital systems, increasing the need for reliable and ...
Fadi Makarem +4 more
wiley +1 more source
A Unified Object Detection Method in Drone View With Degradation‐Aware and Domain Adaptive Modeling
We propose a unified UAV object detection method combining degradation‐aware frequency modeling and domain‐adaptive restoration, enabling consistent detection performance across complex weather conditions. ABSTRACT Existing object detection methods remain severely challenged by adverse weather and domain shifts.
Lixiu Wu, Song Wang
wiley +1 more source
Assessment of a VoIP steganalysis method based on statistical analysis and deep neural network. [PDF]
Moghadasi HA, Dehghani H.
europepmc +1 more source

