3 papers
cs.CV2026
AFFMAE: Scalable Vision Pre-Training for High-Resolution Microscopy Segmentation on Desktop Hardware
David Smerkous, Zian Wang, Behzad Najafian
Self-supervised pretraining has transformed computer vision by enabling data-efficient fine-tuning, yet high-resolution pretraining typically requires server-scale infrastructure,…
cs.LG2026
E2Former-V2: On-the-Fly Equivariant Attention with Linear Activation Memory
Lin Huang, Chengxiang Huang, Ziang Wang +7
Equivariant Graph Neural Networks (EGNNs) have become a widely used approach for modeling 3D atomistic systems. However, mainstream architectures face critical scalability bottlene…
cs.LG2025
HiMoE: Heterogeneity-Informed Mixture-of-Experts for Fair Spatial-Temporal Forecasting
Shaohan Yu, Pan Deng, Yu Zhao +2
Achieving both accurate and consistent predictive performance across spatial nodes is crucial for ensuring the validity and reliability of outcomes in fair spatial-temporal forecas…