AI Paper Atlas

Research map / Video / Video-language understanding

VIR-Bench: Evaluating Geospatial and Temporal Understanding of MLLMs via Travel Video Itinerary Reconstruction

Hao Wang, Eiki Murata, Lingfang Zhang, Ayako Sato, So Fukuda, Ziqi Yin, Wentao Hu, Keisuke Nakao, Yusuke Nakamura, Sebastian Zwirner, Yi-Chia Chen, Hiroyuki Otomo and 2 more

AAAI 2026 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding · 0 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
increase input frames from 64 to 128 (ablation)GPT-4.1 · POI node predictionF1: 39.6 -> 52.9Improved
increase input frames from 64 to 256 (ablation)GPT-4.1 · transition edge predictionF1: 27.6 -> 34.4Improved
increase input frames from 128 to 256 (ablation)GPT-4.1 · inclusion edge predictionF1: 78.8 -> 76.5Mixed
raise reasoning effort from low to high (ablation)o4-mini · transition edge predictionF1: 30.0 -> 43.8Improved
raise reasoning effort from low to high (ablation)o4-mini · prefecture node predictionF1: 86.8 -> 86.4No clear effect
add audio input (ablation)Gemini-2.5-Flash · transition edge predictionF1: 22.3 -> 41.7Improved
add audio input (ablation)Gemini-2.5-Flash · inclusion edge predictionF1: 82.6 -> 78.5Hurt
add audio input (ablation)Gemini-2.5-Flash · prefecture node predictionF1: 82.5 -> 85.1Improved
scale Qwen2.5-VL from 7B to 72BQwen2.5-VL · transition edge predictionF1: 1.3 -> 20.8Improved
scale LLaVA-Video from 7B to 72BLLaVA-Video · POI node predictionF1: 6.0 -> 7.8No clear effect
give agent both POI list and video instead of POI onlytravel-planning agent (Gemini-2.5-Pro) · Attraction (1-5)mean score: 3.58 -> 3.73Improved
use video only instead of POI list for planning agenttravel-planning agent (Gemini-2.5-Pro) · Alignment with videoshare of plans unrelated: 31% completely unrelatedHurt

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