GST-Bench：視覺語言模型能否從影片中發展全域空間感知？

摘要

空間智能對具身智能體至關重要，然而現有基準多專注於單一或少量視角下的局部空間感知，忽略了在連續、長期的視覺串流中進行全域空間感知的能力。為解決此限制，我們提出全域-空間-時間基準（GST-Bench），一個針對影片理解中全域空間智能的視覺問答（VQA）基準，其問題皆由人工驗證，源自 6,790 分鐘的合成生成影片。此基準要求模型能從輸入影片中未出現的新穎視角進行精確的空間推論，並將自我中心觀察對應至全域俯視圖。我們對 22 個先進視覺語言模型（VLM）的全面評估，揭示了模型與人類之間的顯著差距：最強的零樣本模型僅達到 42.68，遠低於人類的 79.08。為探究此差距的成因，我們建構了 GST-Bench-Local，發現模型儘管在相同的任務形式下具有強大的局部空間理解能力，仍無法將長期的觀察結果整合為全域一致的場景表徵。我們進一步提供 GST-Train，一個專為全域空間推理設計的資料集，作為輔助資源，以促進未來對此挑戰的研究。

English

Spatial intelligence is fundamental to embodied agents, yet existing benchmarks focus on local spatial perception from single or few viewpoints, overlooking global spatial awareness over continuous, long-horizon visual streams. To address this limitation, we introduce the Global-Spatial-Temporal Benchmark (GST-Bench), a VQA benchmark for global spatial intelligence in video understanding, comprising human-verified questions derived from 6,790 minutes of synthetically generated video. It requires models to perform accurate spatial inference from novel viewpoints unseen in the input video and to map egocentric observations onto global top-down images. A comprehensive evaluation of 22 state-of-the-art VLMs exposes a striking gap between models and humans: the strongest zero-shot model attains only 42.68, far below the human score of 79.08. To probe the cause of this gap, we construct GST-Bench-Local and find that models, despite strong local spatial understanding under the same task formulation, still fail to consolidate long-horizon observations into a globally consistent scene representation. We further provide GST-Train, a dataset for global spatial reasoning, as a complementary resource to facilitate future research on this challenge.