Frequency-based View Selection in Gaussian Splatting Reconstruction
This addresses the challenge of efficient 3D reconstruction in robotics perception, though it appears incremental as it builds on existing Gaussian Splatting methods.
The paper tackles the problem of active view selection for 3D Gaussian Splatting reconstructions to minimize input images, achieving state-of-the-art results by ranking views in the frequency domain to estimate information gain without ground truth data.
Three-dimensional reconstruction is a fundamental problem in robotics perception. We examine the problem of active view selection to perform 3D Gaussian Splatting reconstructions with as few input images as possible. Although 3D Gaussian Splatting has made significant progress in image rendering and 3D reconstruction, the quality of the reconstruction is strongly impacted by the selection of 2D images and the estimation of camera poses through Structure-from-Motion (SfM) algorithms. Current methods to select views that rely on uncertainties from occlusions, depth ambiguities, or neural network predictions directly are insufficient to handle the issue and struggle to generalize to new scenes. By ranking the potential views in the frequency domain, we are able to effectively estimate the potential information gain of new viewpoints without ground truth data. By overcoming current constraints on model architecture and efficacy, our method achieves state-of-the-art results in view selection, demonstrating its potential for efficient image-based 3D reconstruction.