WarpRF: Multi-View Consistency for Training-Free Uncertainty Quantification and Applications in Radiance Fields
This provides a training-free uncertainty quantification framework for radiance fields, benefiting applications in 3D reconstruction and view synthesis, though it is incremental as it builds on existing consistency assumptions.
The paper tackles the problem of quantifying uncertainty in radiance fields without requiring training, by using multi-view consistency through backward warping, and achieves superior performance in tasks like active view selection and active mapping compared to existing methods.
We introduce WarpRF, a training-free general-purpose framework for quantifying the uncertainty of radiance fields. Built upon the assumption that photometric and geometric consistency should hold among images rendered by an accurate model, WarpRF quantifies its underlying uncertainty from an unseen point of view by leveraging backward warping across viewpoints, projecting reliable renderings to the unseen viewpoint and measuring the consistency with images rendered there. WarpRF is simple and inexpensive, does not require any training, and can be applied to any radiance field implementation for free. WarpRF excels at both uncertainty quantification and downstream tasks, e.g., active view selection and active mapping, outperforming any existing method tailored to specific frameworks.