Search papers, labs, and topics across Lattice.
The paper addresses high-precision dichotomous image segmentation (DIS) by leveraging pseudo-depth information from monocular depth estimation to provide semantic understanding and spatial priors. They introduce a Prior of Depth Fusion Network (PDFNet) that fuses RGB and pseudo-depth features and incorporates a depth integrity-prior loss function to enforce depth consistency. The network also includes a fine-grained perception enhancement module for boundary refinement, achieving state-of-the-art performance with significantly fewer parameters than diffusion-based models.
Monocular depth estimation can unlock surprisingly effective and efficient image segmentation by exploiting the "depth integrity-prior"—the observation that foreground objects exhibit more consistent depth values than backgrounds.
High-precision dichotomous image segmentation (DIS) is a task of extracting fine-grained objects from high-resolution images. Existing methods face a dilemma: non-diffusion methods work efficiently but suffer from false or missed detections due to weak semantics and less robust spatial priors; diffusion methods, using strong generative priors, have high accuracy but encounter high computational burdens. As a solution, we find pseudo depth information from monocular depth estimation models can provide essential semantic understanding that quickly reveals spatial differences across target objects and backgrounds. Inspired by this phenomenon, we discover a novel insight we term the depth integrity-prior: in pseudo depth maps, foreground objects consistently convey stable depth values with much lower variances than chaotic background patterns. To exploit such a prior, we propose a Prior of Depth Fusion Network (PDFNet). Specifically, our network establishes multimodal interactive modeling to achieve depth-guided structural perception by deeply fusing RGB and pseudo depth features. We further introduce a novel depth integrity-prior loss to explicitly enforce depth consistency in segmentation results. Additionally, we design a fine-grained perception enhancement module with adaptive patch selection to perform boundary-sensitive detail refinement. Notably, PDFNet achieves state-of-the-art performance with only 94M parameters (<11% of those diffusion-based models), outperforming all non-diffusion methods and surpassing some diffusion methods. Code is provided in the supplementary materials.
