# TriFlow Generates Artist-Quality 3D Mesh Topology via Nearest-Vertex Vector Fields

_Research · published 2026-07-10_

Researchers from the Technical University of Munich, AUDI AG, and the University of Virginia have introduced TriFlow, a generative method for producing compact, artist-like 3D mesh topology directly from signed distance fields (SDFs). The core innovation is a Nearest-Vertex Vector Field (NVF) that reformulates discrete topology prediction as a piecewise-continuous field problem, which a latent flow-matching model then synthesizes conditioned on the input geometry. A constrained quadric error metric simplification step extracts the final coherent mesh.

Compared to state-of-the-art learning-based approaches, TriFlow achieves 90% lower Chamfer Distance and an 8× speedup, while also supporting level-of-detail control to target different polygon budgets from the same input. The work is accepted to ECCV 2026 and represents a meaningful step toward automating the retopology stage of 3D asset production pipelines.

## Sources
- [derkleineli.github.io](https://derkleineli.github.io/triflow/)
---
Canonical: https://genbuzz.news/posts/triflow-generates-artist-quality-3d-mesh-topology-via-nearest-vertex-vector-fields