AI Reads Well Logs to Map Hidden Volcanic Oil Reservoirs With 98.79% Accuracy
Researchers in China converted well logs into images and used an attention-enhanced convolutional neural network to identify volcanic lithologies in ...
Researchers in China converted well logs into images and used an attention-enhanced convolutional neural network to identify volcanic lithologies in ...
Researchers have built a three-layer AI framework for 5G mobile learning that cuts device energy use by up to 27 ...
An AI tool called ChromAgeNet identifies aging in blood stem cells by detecting subtle age-related patterns in the 3D organization ...
A 3D multi-scale convolutional neural network developed in Taiwan accurately segments and quantifies acute ischemic stroke lesions on diffusion-weighted MRI ...
A 0.15-million-parameter attention-equipped neural network classifies nine wafer-map defect types with 99.88 percent accuracy on a CPU, combining latent-space augmentation ...
A dual-track deep learning network combining Swin Transformer global context with convolutional feature pyramids classifies pear leaf diseases with 92.69% ...
Chinese researchers found that the classic Short-Time Fourier Transform outperforms wavelet and Hilbert–Huang methods when convolutional neural networks classify raw, ...
A new deep learning network grades the gain quality of automated breast ultrasound images with 95.49 percent accuracy, outperforming standard ...
A convolutional neural network trained on rigorous European in situ radiometry now retrieves 13 phytoplankton pigments from Sentinel-3 OLCI, with ...
Researchers have combined a novel 4D sinusoidal chaotic map with an untrained deep network that expands a 4-by-4 seed into ...
© 2026 Scienmag - Science Magazine
© 2026 Scienmag - Science Magazine