Results 31 to 40 of about 82,604 (262)
Leveraging logit uncertainty for better knowledge distillation
Knowledge distillation improves student model performance. However, using a larger teacher model does not necessarily result in better distillation gains due to significant architecture and output gaps with smaller student networks. To address this issue,
Zhen Guo +3 more
doaj +1 more source
Multi-Level Feature Dynamic Fusion Neural Radiance Fields for Audio-Driven Talking Head Generation
Audio-driven cross-modal talking head generation has experienced significant advancement in the last several years, and it aims to generate a talking head video that corresponds to a given audio sequence.
Wenchao Song +4 more
doaj +1 more source
Septin 9 polybasic domains couple phosphoinositide‐rich membrane binding to centrosome positioning, Golgi organization, and microtubule acetylation to control epithelial polarity. Their loss disrupts this axis, causing centrosome mispositioning, Golgi fragmentation, reduced microtubule acetylation, and polarity inversion via upregulation of the ...
Ting ting Cai +4 more
wiley +1 more source
From mice to humans—divergent strategies for intestinal homeostasis and regeneration
Recent advances such as organoid genome editing, xenotransplantation, imaging, and whole‐genome sequencing have enabled direct studies of human intestinal stem cells (ISCs). These studies reveal species‐specific features, including slower ISC proliferation, distinct injury responses, slower somatic mutation accumulation in humans, and an inverse ...
Keiko Ishikawa +2 more
wiley +1 more source
DBDST-Net: Dual-Branch Decoupled Image Style Transfer Network
The image style transfer task aims to apply the style characteristics of a reference image to a content image, generating a new stylized result. While many existing methods focus on designing feature transfer modules and have achieved promising results ...
Na Su +4 more
doaj +1 more source
DRNet: A Deep Neural Network With Multi-Layer Residual Blocks Improves Image Denoising
Recently, with the broad applications of deep learning technology in image denoising, many deep neural networks based on the residual block (ResBlock) have been proposed to improve the performance.
Jiahong Zhang +4 more
doaj +1 more source
Structural insights and therapeutic targets in Acinetobacter baumannii capsule biosynthesis
Hypervirulent KL49 A. baumannii's capsular polysaccharide contains the nonulosonic acid 8‐epi‐Leg5,7Ac2, synthesized by epimerization via ElaA, ElaB, and ElaC. Crystal structures of ElaA, ElaB, and ElaC reveal their role in CMP‐Leg5,7Ac2 synthesis and regioselective C8 epimerization.
Woo Cheol Lee +7 more
wiley +1 more source
Sentiment analysis is pivotal in advancing human–computer interaction (HCI) systems as it enables emotionally intelligent responses. While existing models show potential for HCI applications, current conversational datasets exhibit critical limitations ...
Yao Fu +4 more
doaj +1 more source
Modelling stem cell differentiation related processes—A practical overview for biologists
Stem cell differentiation is complex and difficult to control experimentally. This review introduces suitable computational modelling approaches that can support stem cell research, from mechanistic ODE and abstract models to multiscale and deep learning methods.
Ricco Zeegelaar +4 more
wiley +1 more source
An epithelial GPR35 isoform supports tumor‐associated transcriptional and metabolic phenotypes
GPR35 generates two functionally distinct isoforms with previously unresolved roles. GPR35‐short mediates immune‐cell chemotaxis, while GPR35‐long is enriched in colorectal cancer epithelium, where it supports increased metabolism, proliferation, and tumor‐associated transcriptional programs.
Jørgen D. Rønneberg +14 more
wiley +1 more source

