A three-year benchmark in Indonesia's Sunda Strait shows deep learning models, especially a compact direct multi-step CNN-GRU, deliver the most...

A foundation model-based AI system called WeedNet identifies 1,593 weed species with 91.02 percent accuracy and reaches 97.38 percent accuracy...

A multidimensional computer vision framework measuring aspect ratio, ink coverage, and fractal boundary complexity finds that the disputed Song Dynasty...

Researchers used natural language processing on six years of interviews to show that a large UK research consortium needed nearly...

A new deep-learning method called VIMA converts tissue patches into numerical fingerprints and uses rigorous statistics to uncover disease-associated spatial...

Researchers have developed CytoVI, an open-source deep generative model that integrates, denoises and imputes antibody-based single-cell data from flow cytometry,...

Researchers at Institute of Science Tokyo have developed ITCD, a two-stage diffusion model with trajectory constraints that fills missing values...

A controlled head-to-head study shows Vision Transformers beat convolutional networks on real X-ray diffraction data once fine-tuned, while CNNs remain...

A localized convolutional neural network trained on ASCAT satellite radar data outperforms the operational soil moisture product and, during heavy...

Researchers trained XGBoost machine learning models on automated finite element simulations to predict how special threaded connections respond to combined...