Results 201 to 210 of about 2,004,297 (387)
Neural Activation Constellations: Unsupervised Part Model Discovery with Convolutional Networks [PDF]
Marcel Simon, Erik Rodner
openalex +1 more source
This study presents a wave propagation‐based optimization strategy for asymmetric double‐layer Au gratings to improve infrared polarization selectivity. A logic‐based method efficiently determines single‐wavelength structures through analytical modeling, while a learning‐based approach utilizing an explainable neural network enables broadband ...
Ryuna Kang+4 more
wiley +1 more source
Convolutional Neural Networks at Constrained Time Cost
Kaiming He, Jian Sun
openalex +2 more sources
Accurately predicting protein structure is of great significance in biological research. LightRoseTTA, a light‐weight deep graph network, to achieve prediction for proteins is presented. Notably, three highlights are possessed by LightRoseTTA: i) high‐accurate structure prediction for proteins; ii) high‐efficient training and inference; and iii) low ...
Xudong Wang+7 more
wiley +1 more source
Mitochondria across the entire neuromuscular system have been comprehensively reconstructed at different developmental stages using 3D electron microscopy. Fundamental structural principles related to synaptic function are preserved across development, and these morphologies are adapted to ensure effective neural circuit function.
J. Alexander Bae+9 more
wiley +1 more source
Fully Convolutional Neural Networks for Crowd Segmentation
Kai Kang, Xiaogang Wang
openalex +2 more sources
Epileptiform Activity and Seizure Risk Follow Long‐Term Non‐Linear Attractor Dynamics
This study leverages the HAVOK framework to model long‐term, nonlinear attractor dynamics underlying epileptiform activity and seizure risk in epilepsy patients. By identifying key forcing mechanisms driving chaotic transitions, the findings improve seizure risk forecasting over multi‐day cycles and provide a pathway for personalized, data‐driven ...
Richard E Rosch+4 more
wiley +1 more source
scHeteroNet) is a novel graph neural network model that explicitly addresses heterophily in single‐cell sequencing data enabling accurate cell type annotation and novel cell type discovery. By leveraging heterophily‐aware message passing and novelty propagation mechanisms, scHeteroNet achieves superior performance in both cell annotation and detection ...
Jiacheng Liu+7 more
wiley +1 more source
A Knowledge‐Guided Graph Learning Approach Bridging Phenotype‐ and Target‐Based Drug Discovery
Knowledge‐Guided Drug Relational Predictor (KGDRP), a graph representation learning approach, effectively integrates multiple omics data, including biological network data, gene expression data, and sequence data that incorporates chemical molecular structures.
Qing Ye+10 more
wiley +1 more source