HADS-Net fuses physics-informed texture features from EfficientNet-B3 with Sobel boundary features via cross-attention to reach 96.58% accuracy and 0.9978 macro ROC-AUC on the BUSI breast ultrasound dataset.
Attention is all you need,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
HADS-Net:A Hybrid Attention-Augmented Dual-Stream Network with Physics-Informed Augmentation for Breast Ultrasound Image Classification
HADS-Net fuses physics-informed texture features from EfficientNet-B3 with Sobel boundary features via cross-attention to reach 96.58% accuracy and 0.9978 macro ROC-AUC on the BUSI breast ultrasound dataset.