ICON: Incremental CONfidence for Joint Pose and Radiance Field Optimization
This addresses the challenge of pose initialization in NeRF training for 3D reconstruction and view synthesis, offering a more flexible approach for applications like robotics or AR, though it builds incrementally on prior work by refining pose estimation methods.
The paper tackles the problem of training Neural Radiance Fields (NeRF) for novel view synthesis without requiring accurate initial camera poses, by introducing ICON, an optimization procedure that uses smooth camera motion assumptions and adaptive confidence measures, achieving superior performance on CO3D and HO3D datasets compared to methods using Structure-from-Motion poses.
Neural Radiance Fields (NeRF) exhibit remarkable performance for Novel View Synthesis (NVS) given a set of 2D images. However, NeRF training requires accurate camera pose for each input view, typically obtained by Structure-from-Motion (SfM) pipelines. Recent works have attempted to relax this constraint, but they still often rely on decent initial poses which they can refine. Here we aim at removing the requirement for pose initialization. We present Incremental CONfidence (ICON), an optimization procedure for training NeRFs from 2D video frames. ICON only assumes smooth camera motion to estimate initial guess for poses. Further, ICON introduces ``confidence": an adaptive measure of model quality used to dynamically reweight gradients. ICON relies on high-confidence poses to learn NeRF, and high-confidence 3D structure (as encoded by NeRF) to learn poses. We show that ICON, without prior pose initialization, achieves superior performance in both CO3D and HO3D versus methods which use SfM pose.