MMVIAD - Defect Classification: leaderboard
MMVIAD (Zhao et al., 2026) is rendered 2-second inspection videos that sweep about 120 degrees around Anomaly-ShapeNet objects (48 categories, six structural defect types), asking whether the object is anomalous, which defect it has, what the object is and when the defect is visible. This board is the defect-type question on the Standard test split, scored by accuracy in percent; higher is better.
Metric: Accuracy (%; MMVIAD-Standard test split, 1,101 rendered 2-second inspection clips over the same 48 object categories as training, 32 frames, which of six defect types or none). Source: arxiv.org. Saturation forecast: Around October 2027. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 41.6 | #91 |
| 2 | GPT-5.4 | 37.9 | #121 |
| 3 | GPT-5.4 Mini | 32.6 | #389 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=mmviad-defect-classification · How It Works · Data refreshed daily, snapshot 2026-11-12.