Tag: deep learning

Tautomerism occupies a curious position in chemistry: it is conceptually simple, involving nothing more than the relocation of a single ...

Researchers in Trieste embedded anatomical rules as logical constraints in a transformer segmentation model, boosting accuracy and morphological coherence in ...

A new deep learning model combines diffusion-style state propagation with Mamba-based sequence modeling and window attention to grade ulcerative colitis ...

A new calibration-aware Vision Transformer framework classifies brain tumor MRI scans with over 98 percent accuracy while flagging uncertain cases ...

A sweeping review finds AI delivers impressive agricultural gains in irrigation, soil management, and pest detection, but warns that spatial ...

A German research team has shown that a GLM-based residualization method removes age confounding from MRI-based Alzheimer's classifiers far more ...

A new iScience study introduces CSCC-PAG, a deep learning framework that combines domain-specific pre-training with self-supervised domain adaptation to improve ...

A transformer-based deep-learning pipeline called ProFormer classifies single-cell and plasma proteomes directly from raw mass-spectrometry fingerprints, outperforming CNNs and enabling ...

Plant pathologists have long faced a stubborn diagnostic problem: a single leaf can carry several diseases at once, each with ...

A new detection framework combines hybrid attention, a slim dynamic head, and a novel loss function to spot crop diseases ...

Deep learning models that promise to weigh pastures from photographs may be far less trustworthy than their published accuracy figures ...

A new deep clustering method, scDAEC, combines two autoencoders and multi-head attention to untangle noisy single-cell RNA sequencing data, outperforming ...

A new systematic review of 150 studies argues that AI-driven cybersecurity for the Industrial Internet of Things has reached impressive ...

A new deep learning network called MALCNet reconstructs missing MRI information in the latent space, achieving statistically significant gains in ...

A deep learning model that fuses PET imaging features with clinical data predicted breast cancer molecular subtypes non-invasively, reaching an ...

Researchers trained a deep reinforcement learning agent to plan collision-free paths for a fuel-saving free-floating space manipulator, even with noisy ...

A major new review maps how artificial intelligence is transforming biological age prediction, from epigenetic clocks to multimodal deep learning, ...

Researchers in Thailand have developed an explainable deep learning system that diagnoses five causes of vaginal discharge from smartphone photos ...

Researchers in Xi'an have trained a cascaded generative adversarial network to synthesize 18F-FDG PET images from 18F-octreotide scans, achieving near-perfect ...

A new review maps how artificial intelligence is transforming RNA-targeted drug discovery across both small molecules and oligonucleotide therapeutics.