AI Paper Atlas

Research map / Video / Video-language understanding

Pistachio: Towards Synthetic, Balanced, and Long-Form Video Anomaly Benchmarks

jie li, Hongyi Cai, Mingkang Dong, Muxin Pu, Shan You, Fei Wang, Tao Huang

ECCV 2026 · Video › Video-language understanding · approach: Event and anomaly reasoning in video · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
train on Pistachio plus 10% of UCF-Crime instead of full UCF-Crime only (ablation)CLIP-TSA · UCF-Crime testAUC / AP: 82.56/21.62 -> 83.96/24.25Improved
increase the UCF-Crime share mixed with Pistachio from 10% to 50% (ablation)CLIP-TSA · UCF-Crime testAP: 24.25 -> 33.04Improved
train on Pistachio only instead of UCF-CrimeCLIP-TSA · UCF-Crime testAUC: 82.56 -> 61.58Hurt
train on Pistachio instead of UCF-CrimeCLIP-TSA · XD-Violence testAP: 51.75 -> 62.48Improved
train on Pistachio instead of UCF-CrimeCLIP-TSA · Hailuo testAUC: 68.71 -> 75.89Improved
joint training on all five datasets instead of Pistachio only (ablation)CLIP-TSA · Cosmos testAP: 58.74 -> 87.37Improved
use training-free VADTree instead of trained methodsVADTree · Pistachio VAD testAP: 25.15 AP (72.52 AUC) vs Fed-WSVAD 71.9Hurt
use storyline-based last-frame chaining instead of direct one-shot long generation (ablation)Wan generation pipeline · VLM quality scoreAvg. quality: 2.75 -> 4.17Improved
add multi-step storyline to chained generation (ablation)Wan generation pipeline · VLM quality scoreAvg. quality: 3.45 -> 4.17Improved
use last-frame chaining instead of fixed first-last-frame looping (ablation)Wan generation pipeline · VLM quality scoreAvg. quality: 4.04 -> 4.17Improved
scale VAU model from InternVL3-8B to 14B (ablation)InternVL3 · Pistachio-VAUAvg F1: 29.11 -> 29.39No clear effect
add 4-step distillation and TeaCache to Wan generation (ablation)Wan generation · Generation efficiencytime per video: 349.7s -> 30.0sImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy