AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

STI-Bench: Are MLLMs Ready for Precise Spatial-Temporal World Understanding?

Yun Li, Yiming Zhang, Tao Lin, Xiangrui Liu, Wenxiao Cai, Zheng Liu, Bo Zhao

ICCV 2025 · Vision-language and multimodal › Multimodal LLMs · approach: Spatial reasoning in VLMs · 61 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
use Gemini-2.5-Pro instead of GPT-4oMLLM · STI-Benchaccuracy: 34.8 -> 41.4Improved
use larger MLLM (Qwen2.5-VL-72B) instead of MiniCPM-V-2.6open-source MLLM · STI-Benchaccuracy: 26.9 -> 40.7Improved
evaluate outdoor instead of desktop scenesGemini-2.5-Pro · STI-Benchaccuracy: 35.8 -> 48.7Improved
evaluate outdoor instead of desktop scenesQwen2.5-VL-72B · STI-Benchaccuracy: 33.5 -> 50.6Improved
precise quantitative task (dimensional measurement) vs pose estimationGemini-2.5-Pro · STI-Benchaccuracy: 38.7 vs 50.4Hurt
displacement and path length vs spatial relationGemini-2.5-Pro · STI-Benchaccuracy: 33.9 vs 53.8Hurt
open-source Qwen2.5-VL-72B vs proprietary Gemini-2.5-ProMLLM · STI-Benchaccuracy: 40.7 vs 41.4No clear effect
ego-centric orientation: Gemini-2.0-Flash vs Gemini-2.5-ProGemini · STI-Bench Ego Orient.accuracy: 10.8 -> 52.5Improved
best model vs random baselineGemini-2.5-Pro · STI-Benchaccuracy: 20 -> 41.4Improved

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