FullCircle: Effortless 3D Reconstruction from Casual 360$^\circ$ Captures

Abstract

Radiance fields have emerged as powerful tools for 3D scene reconstruction. However, casual capture remains challenging due to the narrow field of view of perspective cameras, which limits viewpoint coverage and feature correspondences necessary for reliable camera calibration and reconstruction. While commercially available 360$^\circ$ cameras offer significantly broader coverage than perspective cameras for the same capture effort, existing 360$^\circ$ reconstruction methods require special capture protocols and pre-processing steps that undermine the promise of radiance fields: effortless workflows to capture and reconstruct 3D scenes. We propose a practical pipeline for reconstructing 3D scenes directly from raw 360$^\circ$ camera captures. We require no special capture protocols or pre-processing, and exhibit robustness to a prevalent source of reconstruction errors: the human operator that is visible in all 360$^\circ$ imagery. To facilitate evaluation, we introduce a multi-tiered dataset of scenes captured as raw dual-fisheye images, establishing a benchmark for robust casual 360$^\circ$ reconstruction. Our method significantly outperforms not only vanilla 3DGS for 360$^\circ$ cameras but also robust perspective baselines when perspective cameras are simulated from the same capture, demonstrating the advantages of 360$^\circ$ capture for casual reconstruction. Additional results are available at: https://theialab.github.io/fullcircle

Figures (12)

• Figure 1: Teaser -- With 360$^\circ$ cameras we can recover 3D scenes far more effectively than with traditional perspective cameras, resulting in substantial gains in novel view synthesis (left: perspective vs. 360$^\circ$). However, casual 360$^\circ$ captures unavoidably include the camera operator, which degrades reconstruction quality if left unaddressed (e.g. 3DGRT 3dgrt). Our FullCircle method overcomes this challenge, enabling fast, high-quality scene capture using 360$^\circ$ cameras.
• Figure 2: Stitching artifacts -- Equirectangular inputs contain stitching artifacts (top left) that lead to noisy edges and reduced PSNR in the reconstruction (bottom left). Our method, trained on raw fisheye images, avoids these artifacts (right).
• Figure 3: Boundary masking -- Unmasked black boundary pixels contaminate the reconstruction (left). Minimal masking (pink) leaves blue artifacts from edge distortion and color aberration (middle). Our dilated mask prevents artifacts (right).
• Figure 4: Capturer mask estimation -- Our masking pipeline first detects the capturer using YOLOv8 yolo and SAMv2 sam (red masks). These initial masks roughly localize the capturer but include missing or incorrect regions that can degrade reconstruction. We then re-center the omni images on the detected capturer, render synthetic fisheyes, and re-run SAMv2 with a center-point prompt to obtain refined, temporally consistent masks (purple masks), which are mapped back to the original dual-fisheye inputs.
• Figure 5: Dual-fisheye vs. perspective capture -- Using dual-fisheye images for both calibration and reconstruction ( – ) yields higher-quality novel views and stable geometry than when reconstruction ( – ), or both calibration and reconstruction ( – ), relies on perspective images. This advantage becomes most apparent when moving beyond the in-distribution training trajectory (purple cameras) toward out-of-distribution test views (yellow camera), where ( – ) and ( – ) exhibit a gradual degradation in reconstruction quality.