Triangle Splatting+: Differentiable Rendering with Opaque Triangles

2025-09-29

AI Summary

3D Gaussian Splatting (3DGS) suffers from incompatibility with standard graphics engines (e.g., VR headsets, real-time rendering pipelines), necessitating post-hoc mesh conversion that incurs quality degradation and pipeline complexity. Method: We propose Triangle Splatting+, the first differentiable semi-connected triangular representation framework. It employs shared-vertex triangle meshes as the fundamental differentiable primitive, augmented with explicit opacity constraints and end-to-end gradient-based optimization. Contribution/Results: Triangle Splatting+ directly outputs topologically valid, geometrically crisp, and production-ready triangle meshes. On Mip-NeRF360 and Tanks & Temples benchmarks, it achieves state-of-the-art performance for mesh-based novel-view synthesis—surpassing leading splatting methods in visual fidelity, training efficiency, and robustness—while natively enabling downstream applications such as physics simulation.

Technology Category

• Computer Vision: 3D Computer Vision
• Search and Optimization: Sampling/Simulation-based Search
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Security and Privacy: Data transparency and provenance
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Reconstructing 3D scenes and synthesizing novel views has seen rapid progress in recent years. Neural Radiance Fields demonstrated that continuous volumetric radiance fields can achieve high-quality image synthesis, but their long training and rendering times limit practicality. 3D Gaussian Splatting (3DGS) addressed these issues by representing scenes with millions of Gaussians, enabling real-time rendering and fast optimization. However, Gaussian primitives are not natively compatible with the mesh-based pipelines used in VR headsets, and real-time graphics applications. Existing solutions attempt to convert Gaussians into meshes through post-processing or two-stage pipelines, which increases complexity and degrades visual quality. In this work, we introduce Triangle Splatting+, which directly optimizes triangles, the fundamental primitive of computer graphics, within a differentiable splatting framework. We formulate triangle parametrization to enable connectivity through shared vertices, and we design a training strategy that enforces opaque triangles. The final output is immediately usable in standard graphics engines without post-processing. Experiments on the Mip-NeRF360 and Tanks & Temples datasets show that Triangle Splatting+achieves state-of-the-art performance in mesh-based novel view synthesis. Our method surpasses prior splatting approaches in visual fidelity while remaining efficient and fast to training. Moreover, the resulting semi-connected meshes support downstream applications such as physics-based simulation or interactive walkthroughs. The project page is https://trianglesplatting2.github.io/trianglesplatting2/.

Problem

• Optimizes triangles directly for differentiable rendering
• Enables mesh compatibility with standard graphics engines
• Achieves real-time novel view synthesis without post-processing

Innovation

• Directly optimizes triangles in differentiable splatting framework
• Enforces opaque triangles through specialized training strategy
• Produces semi-connected meshes compatible with standard graphics engines