Results 191 to 200 of about 25,807,331 (282)
KDLM: Lightweight Brain Tumor Segmentation via Knowledge Distillation
Advanced Intelligent Systems, EarlyView.A lightweight student network is designed, which is based on multiscale and multilevel feature fusion and combined with the residual channel attention mechanism to achieve efficient feature extraction and fusion with very few parameters. A dual‐teacher collaborative knowledge distillation framework is proposed.
Baotian Li +4 more
wiley +1 more source
A Flexible and Energy‐Efficient Compute‐in‐Memory Accelerator for Kolmogorov–Arnold Networks
Advanced Intelligent Systems, EarlyView.This article presents KA‐CIM, a compute‐in‐memory accelerator for Kolmogorov–Arnold Networks (KANs). It enables flexible and efficient computation of arbitrary nonlinear functions through cross‐layer co‐optimization from algorithm to device. KA‐CIM surpasses CPU, ASIC, VMM‐CIM, and prior KAN accelerators by 1–3 orders of magnitude in energy‐delay ...
Chirag Sudarshan +6 more
wiley +1 more source
Advanced Intelligent Systems, EarlyView.Four decades of retinal vessel segmentation research (1982–2025) are synthesized, spanning classical image processing, machine learning, and deep learning paradigms. A meta‐analysis of 428 studies establishes a unified taxonomy and highlights performance trends, generalization capabilities, and clinical relevance.
Avinash Bansal +6 more
wiley +1 more source
Methods for Setting Device Specifications for Analog In‐Memory Computing Inference
Advanced Intelligent Systems, EarlyView.Non‐volatile memories (NVMs) are being developed for analog in‐memory computing for energy‐efficient, high‐speed deep learning inference. As technology is moving to industry adoption, a method to define required NVM specifications is critical for improving performance and reducing manufacturing cost.
Zhenyu Wu +3 more
wiley +1 more source
Advanced Intelligent Systems, EarlyView.This work proposes RT‐DETR‐DA, an enhanced real‐time detection framework for identifying distracted driving in complex, real‐world environments. The model introduces a dynamic sparse gating multiscale attention module and an attention‐guided dual‐path fusion module to strengthen multiscale perception and cross‐layer feature interaction.
Yi Liu +4 more
wiley +1 more source
Advanced Intelligent Systems, EarlyView.This work presents a comprehensive framework bridging device fabrication, modeling, and system‐level simulation for an indium‐gallium‐zinc‐oxide (IGZO) charge‐trap synaptic transistor‐based neuromorphic system. By developing a precise SPICE model derived from fabricated IGZO synaptic transistors, the study incorporates parasitic RC loads into array ...
Yumin Yun +5 more
wiley +1 more source
American Journal of Agricultural Economics, EarlyView.Abstract This study develops and empirically estimates a structural framework to decompose the causal pathways of multilevel behavioral interventions targeting adolescent health behaviors. We apply this framework to the Kids SIPsmartER (KSS) program, a 6‐month, school‐based intervention evaluated through a clustered randomized controlled trial in rural
Naveen Abedin +5 more
wiley +1 more source
American Journal of Community Psychology, EarlyView.Abstract Despite the heightened mental health challenges amid rising Anti‐Asian sentiment, Asian Americans have significantly underutilized mental health services, a trend that persisted even before the COVID‐19 pandemic. Although considerable efforts have been made to understand how various factors are related to mental health service use in this ...
Michael Park +6 more
wiley +1 more source