Depth State Space Model for Light Field Depth Estimation via Text-Similar Representation

Light field (LF) technology captures both spatial and angular information of the real world, enabling accurate depth estimation. Cost volume-based methods mostly consider LF depth estimation as a shift-matching process, which fail to efficiently establish the relationship among different viewpoints. State Space Model (SSM) has shown strong capabilities in long-sequence modeling, providing a powerful mechanism to capture viewpoints associations. In this paper, we observe that LF depth estimation can be viewed as state transition and then propose a text-similar representation based on the distribution of pixel values across different viewpoints, which is able to detect occluded and discontinuous regions. Furthermore, to extract the potential depth features, we represent it as Depth State Space Model (DSSM), leveraging the state transition mechanism of SSM to capture spatial, angular and structural characteristics in complex regions. Based on the proposed DSSM, we develop DSS-Net for depth estimation. Experiments demonstrate that our approach achieves state-of-the-art performance, with significant improvements in occluded and discontinuous regions, highlighting its effectiveness in addressing the complexities of LF depth estimation.