Research map / 3D vision / 3D generation

FoundSurface: Feedforward Ray-Based 3D Scene Surface Reconstruction from Unposed Images

NeurIPS 2026 · 3D vision › 3D generation · approach: Learning 3D reconstruction from images

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes FoundSurface, a feed-forward pipeline that recovers continuous 3D surfaces from unposed images by combining pointmap-based and ray-based representations.

Problem

Recovering accurate, continuous 3D surfaces from unposed images; existing feed-forward pointmap methods need no camera poses but produce discrete, sparse outputs.

Approach

Integrates pointmap-based and ray-based representations in a single generalizable feed-forward pipeline so that continuous, high-fidelity surfaces are obtained directly from unposed images without per-scene optimization.

Closest papers by meaning

• World Tracing: Pixel-Aligned Geometry Beyond the Visible (NeurIPS 2026 · similarity 0.88)
• SPAR3D: Stable Point-Aware Reconstruction of 3D Objects from Single Images (CVPR 2025 · similarity 0.88)
• UniSDF: Unifying Neural Representations for High-Fidelity 3D Reconstruction of Complex Scenes with Reflections (NeurIPS 2024 · similarity 0.88)
• RaySt3R: Predicting Novel Depth Maps for Zero-Shot Object Completion (NeurIPS 2025 · similarity 0.87)
• GeoSVR: Taming Sparse Voxels for Geometrically Accurate Surface Reconstruction (NeurIPS 2025 · similarity 0.87)
• SurfelSplat: Learning Efficient and Generalizable Gaussian Surfel Representations for Sparse-View Surface Reconstruction (NeurIPS 2025 · similarity 0.87)