Lightweight AI Network Reconstructs High-Resolution Images Under One Million Parameters
A new hierarchical feature attention network reconstructs high-resolution images from low-resolution inputs while keeping the model below one million parameters.
Tag: Neural Networks
Deep learning review maps how AI forecasts renewable energy
A systematic review in Neural Computing and Applications surveys deep learning and hybrid neural models used to forecast renewable energy ...
Greedy Deep Kernel Method Outperforms Neural Networks in Approximation Tests
Researchers combined greedy kernel interpolation with trainable deep kernels and found the resulting models matched or beat neural networks in ...
AI Shows Promise for Predicting Gum Disease and Tooth Loss, but Is Not Ready for the Clinic
A scoping review of eight studies finds AI models predicting periodontal disease progression and tooth loss perform moderately to highly ...
Kernel methods join the Deep Ritz method with new error guarantees
A new open-access paper proves a priori error estimates for kernel methods used as ansatz functions within the Deep Ritz ...
Machine learning predicts high-speed train ground vibration in seconds
A neural network trained on semi-analytical simulations and field measurements predicted elevated-track ground vibration with a maximum reported relative deviation ...
Runtime monitor ensemble proposed for autonomous-driving perception safety
A conceptual study maps three complementary runtime monitors onto ISO 21448 safety activities for AI-based autonomous-driving perception systems, but does ...
SCSNet Improves Early Rumor Detection Across Unseen Events
SCSNet weighted early replies by estimated contribution and used gated stage-wise updates to improve rumor detection within 10 to 30 ...
MCGE-Net joins image, text and labels to enhance imaging
A controlled study reports that MCGE-Net enhanced images using reference images, text and pixel labels, outperforming comparison methods across medical, ...
XPAR-Net: cross-paradigm alignment and reconstruction network for high-fidelity lens-free on-chip microscopy
XPAR-Net improved average reconstruction quality over HoloFormer on an 11-category in-distribution test set while using 8.8% of its trainable parameters ...
Explainable Volterra Networks Bring Transparent Nonlinear Dynamics to Time Series Forecasting
A new neural architecture converts the classical Volterra series into learnable coefficient tensors, recovering known chaotic dynamics, matching transformer physics, ...
Adaptive Temperature Distillation Shrinks Multilingual AI Models for Classroom Use
A new distillation framework that adjusts its temperature per language and sample compresses a 560-million-parameter multilingual model into a 66-million-parameter ...
COSMOS Learns Neural Representations as Gauge Dynamics on the Grassmannian
The COSMOS framework trains orthonormal subspace frames with conjugation-based gauge-style updates, holding geometric integrity at machine precision and matching a ...
Physics Meets Machine Learning to Predict the Full Seismic Response of Steel–Concrete Composite Structures
A new physics-informed machine learning framework embeds hysteretic mechanical laws into neural network training, reproducing full seismic load–deformation and energy ...
Hybrid Deep Learning Model Boosts Prediction of Slope Stability Safety Factor
A hybrid deep learning framework that combines unsupervised feature extraction with supervised regression outperformed a standard neural network and conventional ...
Feed-forward neural basis model delivers most accurate held-out temperature predictions for Schottky diode curves
A leave-one-temperature-out comparison of ten models found that the feed-forward N-BEATS architecture gave the most accurate predictions of Schottky diode ...
Hybrid AI model finds balanced mix for marine coral concrete
A response-surface-guided neural network predicted a balanced silica-fume and nano-silica formulation for coral aggregate concrete, reducing prediction error without additional ...
AMG-based real-time prosthetic hand control: Cross-subject evaluation and embedded implementation
A six-channel acoustomyography bracelet recognized five hand gestures with 95.3% mean accuracy in an online test on a microcontroller-controlled 5-DOF ...
Transformer Improves Critical Heat-Flux Prediction Under Non-Uniform Reactor Heating
A causal transformer reduced critical heat-flux prediction scatter in a held-out benchmark of 109 non-uniform-heating reactor experiments while retaining classical ...
KAN surrogates closely reproduce LPA perfectionism profiles in Turkish adolescents
A comparative study found that Kolmogorov–Arnold Network surrogates closely reproduced LPA perfectionism-profile assignments in 1,385 Turkish adolescents, with EfficientKAN offering ...
POPULAR NEWS
-
Multitask Cambrian Model Improves Protein Melting Temperature Prediction
-
Vasopressors Exert Divergent Immunomodulatory Effects Across a Spectrum in Septic Shock
-
Study Identifies Key Drivers Shaping Salmonid Gill Microbiomes in Aquaculture
-
Chinese Cancer Association issues unified guideline for tumor-related lung injury care