State space duality for multimodal image registration

Arxiv CS.CV (computer vision)

Paper proposes cross-modality feature fusion using Structured State Space Duality (SSD) for multi-modal image registration. SSD method offers better global structural feature extraction and efficiency compared to Transformers.

state-space-modelsmultimodalimage-registration

Arxiv CS.CV (computer vision)·Zhikang LiYan WuXin HuYi DaiMing Li

8 days ago