🤖 AI Summary
To address accuracy degradation in unsupervised monocular depth estimation caused by weak-texture regions and dynamic objects, this paper proposes a self-supervised learning framework integrating geometric priors and differential constraints. Our method introduces (1) dense correspondence priors to enforce explicit geometric consistency; (2) a context-geometry depth consistency loss and a divergence-depth gradient correlation loss that couples optical flow divergence with depth gradients; and (3) a bidirectional calibration mechanism jointly optimizing rigid-flow and optical-flow predictions. By leveraging triangulation-guided depth initialization, differential property modeling, and dual-stream coupled optimization, our approach achieves state-of-the-art performance on KITTI and Make3D benchmarks. It significantly improves depth accuracy in weak-texture regions and enhances local smoothness and structural plausibility in dynamic areas.
📝 Abstract
There has been a recent surge of interest in learning to perceive depth from monocular videos in an unsupervised fashion. A key challenge in this field is achieving robust and accurate depth estimation in challenging scenarios, particularly in regions with weak textures or where dynamic objects are present. This study makes three major contributions by delving deeply into dense correspondence priors to provide existing frameworks with explicit geometric constraints. The first novelty is a contextual-geometric depth consistency loss, which employs depth maps triangulated from dense correspondences based on estimated ego-motion to guide the learning of depth perception from contextual information, since explicitly triangulated depth maps capture accurate relative distances among pixels. The second novelty arises from the observation that there exists an explicit, deducible relationship between optical flow divergence and depth gradient. A differential property correlation loss is, therefore, designed to refine depth estimation with a specific emphasis on local variations. The third novelty is a bidirectional stream co-adjustment strategy that enhances the interaction between rigid and optical flows, encouraging the former towards more accurate correspondence and making the latter more adaptable across various scenarios under the static scene hypotheses. DCPI-Depth, a framework that incorporates all these innovative components and couples two bidirectional and collaborative streams, achieves state-of-the-art performance and generalizability across multiple public datasets, outperforming all existing prior arts. Specifically, it demonstrates accurate depth estimation in texture-less and dynamic regions, and shows more reasonable smoothness.