3D Arena: An Open Platform for Generative 3D Evaluation
Dylan Ebert

TL;DR
3D Arena is a large-scale human preference platform for evaluating generative 3D models, providing reliable rankings and insights into human preferences, surpassing traditional automated metrics.
Contribution
The paper introduces 3D Arena, the largest human preference evaluation platform for generative 3D models, with a new dataset, quality control, and insights into human preferences.
Findings
Gaussian splat outputs outperform meshes and textured models in human preference.
The platform collected over 123,000 votes from more than 8,000 users.
Reliable ELO-based rankings demonstrate the platform's effectiveness.
Abstract
Evaluating Generative 3D models remains challenging due to misalignment between automated metrics and human perception of quality. Current benchmarks rely on image-based metrics that ignore 3D structure or geometric measures that fail to capture perceptual appeal and real-world utility. To address this gap, we present 3D Arena, an open platform for evaluating image-to-3D generation models through large-scale human preference collection using pairwise comparisons. Since launching in June 2024, the platform has collected 123,243 votes from 8,096 users across 19 state-of-the-art models, establishing the largest human preference evaluation for Generative 3D. We contribute the iso3d dataset of 100 evaluation prompts and demonstrate quality control achieving 99.75% user authenticity through statistical fraud detection. Our ELO-based ranking system provides reliable model assessment, with…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
