This paper proposes a novel method for improving the accuracy of supervised monocular depth estimation. We constructed a dual-branch planarguided network which is composed of the previous local planar guidance module and our nonplanar depth prediction...
This paper proposes a novel method for improving the accuracy of supervised monocular depth estimation. We constructed a dual-branch planarguided network which is composed of the previous local planar guidance module and our nonplanar depth prediction module. The outputs of the two sub-networks are aggregated by utilizing the weights computed from the proposed planar-aware guidance module. We collected a real-world indoor dataset, and experimental results demonstrate that the proposed method is effective to improve the depth estimation results.