Dongrui Wu

CoMA-DiT is proposed, a bidirectional cross-modal Diffusion Transformer for latent augmentation that treats paired modalities as sources of mutual generative supervision rather than merely as inputs to be fused.

RAMamba-Net is proposed, a reliability-aware Mamba-based multimodal fusion network for AAD that effectively exploits complementary EEG-EOG information, yielding accuracy gains over unimodal baselines, and is robust to signal perturbation and parameter variation.

Anchoring neural and visual representations can boost brain-to-image retrieval accuracy by over 9% with minimal stimulus repetitions, challenging the reliance on trial averaging.

SCORE achieves a remarkable 53.23% Top-1 accuracy in EEG-to-image retrieval, outperforming existing methods by over 17 percentage points, even without target labels.

STEAM achieves superior EEG decoding performance with a novel hierarchical pre-training approach that enhances model specialization without starting from scratch.