Adapting to the Stream: An Instance-Attention GNN Method for Irregular Multivariate Time Series Data
DynIMTS replaces static graphs with instance-attention that updates edge weights on the fly, delivering SOTA imputation and P12 classification ...
Researchers developed an Ag/Sb2O3/Au memristor array that mimics brain-like computing, performing on-device image feature extraction with low power consumption, promising smarter and faster electric ...
Abstract: Positron range (PR) blurring degrades positron emission tomography (PET) image resolution, particularly for high-energy emitters like 68Ga. We introduce Dual-Input Dynamic Convolution ...
Abstract: The radar reflectivity mosaic image derived from ground-based weather radar observations is crucial for monitoring severe weather events. But radar deployment is non-existent in desert, ...
A research team has developed a new model, PlantIF, that addresses one of the most pressing challenges in agriculture: the ...
Research code for semantic segmentation of 3D plant point clouds using geometric deep learning, developed in collaboration with Oak Ridge National Laboratory.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback