Incremental Joint Learning of Depth, Pose and Implicit Scene Representation on Monocular Camera in Large-scale Scenes

Dense scene reconstruction for photo-realistic view synthesis has various applications, such as VR/AR, autonomous vehicles. However, most existing methods have difficulties in large-scale scenes due to three core challenges: $\mathit{\text{(a)}}\mathit{\text{inaccurate}}\mathit{\text{depth}}\mathit{\text{input.}}$ Accurate depth input is impossible to get in real-world large-scale scenes. $\mathit{\text{(b)}}\mathit{\text{inaccurate}}\mathit{\text{pose}}\mathit{\text{estimation.}}$ Most existing approaches rely on accurate pre-estimated camera poses. $\mathit{\text{(c)}}\mathit{\text{insufficient}}\mathit{\text{scene}}\mathit{\text{representation}}\mathit{\text{capability.}}$ A single global radiance field lacks the capacity to effectively scale to large-scale scenes. To this end, we propose an incremental joint learning framework, which can achieve accurate depth, pose estimation, and large-scale scene reconstruction. A vision transformer-based network is adopted as the backbone to enhance performance in scale information estimation. For pose estimation, a feature-metric bundle adjustment (FBA) method is designed for accurate and robust camera tracking in large-scale scenes. In terms of implicit scene representation, we propose an incremental scene representation method to construct the entire large-scale scene as multiple local radiance fields to enhance the scalability of 3D scene representation. Extended experiments have been conducted to demonstrate the effectiveness and accuracy of our method in depth estimation, pose estimation, and large-scale scene reconstruction.