AutoRecon: Automated 3D Object Discovery and Reconstruction
This addresses the need for automated digital content creation by eliminating manual labor in background removal for 3D reconstruction, though it appears incremental as it builds on existing reconstruction methods.
The authors tackled the problem of automating 3D object reconstruction from multi-view images by developing AutoRecon, which robustly locates and segments foreground objects using self-supervised 2D vision transformer features and reconstructs them with decomposed neural scene representations, achieving effective results on DTU, BlendedMVS, and CO3D-V2 datasets.
A fully automated object reconstruction pipeline is crucial for digital content creation. While the area of 3D reconstruction has witnessed profound developments, the removal of background to obtain a clean object model still relies on different forms of manual labor, such as bounding box labeling, mask annotations, and mesh manipulations. In this paper, we propose a novel framework named AutoRecon for the automated discovery and reconstruction of an object from multi-view images. We demonstrate that foreground objects can be robustly located and segmented from SfM point clouds by leveraging self-supervised 2D vision transformer features. Then, we reconstruct decomposed neural scene representations with dense supervision provided by the decomposed point clouds, resulting in accurate object reconstruction and segmentation. Experiments on the DTU, BlendedMVS and CO3D-V2 datasets demonstrate the effectiveness and robustness of AutoRecon.