Tag: multi-task learning

A new U-Net-based framework uses anatomy-guided, spatially varying regularization to reconstruct chest CT images from sparse-view data with improved performance ...

A unified deep learning framework called MyoSpectrumNet combines pathologic myopia screening, maculopathy grading, refraction estimation and lesion segmentation from heterogeneous ...

A dual-channel machine-learning framework combined sentiment, academic performance and engagement data to predict educational AI adoption, learning gains and personalized ...

FewShot AdaptiveMLC raised macro-F1 by 0.072 on a multi-label Bangla hate speech dataset by fine-tuning task-specific heads for minority labels ...

A single-university study found that jointly predicting semester course sets and grades reduced a Transformer’s mean absolute grade-prediction error by ...

UAC-Net combined contextual attention and multi-scale feature fusion in a shared neural network that classified three brain-tumor types and delineated ...

Researchers at the Vellore Institute of Technology developed three hybrid deep learning models combining convolutional networks, transformers, and clinical metadata ...

A multi-task deep learning system reads infant hip ultrasounds, segments seven anatomical structures, localizes six landmarks, and measures Graf angles ...

Researchers have developed and validated an AI-driven adaptive learning system that predicts student engagement, forecasts academic achievement, and generates explainable ...

Researchers built ViTDrone, a Vision Transformer framework that simultaneously steers a drone through corridors and detects collisions while using Grad-CAM ...