MMVIAD - Anomaly Visible-Time Localization: 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 time interval in which the defect is visible, on the Standard test split, scored by mean temporal IoU in percent; higher is better.
Metric: Mean temporal IoU (%; MMVIAD-Standard test split, predicted versus annotated interval in which the defect is visible, 32 frames). Source: arxiv.org. Saturation forecast: Around September 2027. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 50.5 | #121 |
| 2 | Gemini 3.1 Pro (Preview) | 48.5 | #91 |
| 3 | GPT-5.4 Mini | 46.3 | #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-anomaly-visible-time-localization · How It Works · Data refreshed daily, snapshot 2026-11-12.