Physics Question Scene Graph: Fine-grained Evaluation of Physical Plausibility in Text-to-Video Generation

• URL: http://arxiv.org/abs/2606.25306v1
• Date: Wed, 24 Jun 2026 02:12:54 GMT
• Title: Physics Question Scene Graph: Fine-grained Evaluation of Physical Plausibility in Text-to-Video Generation
• Abstract summary: We introduce Physics Question Scene Graph (PQSG), a hierarchical question-based evaluation pipeline.<n>PQSG evaluates generated videos by checking their faithfulness to a prompt across objects, actions, and adherence to physical laws.<n>We validate PQSG by creating FinePhyEval, a dataset with physics-based prompts and corresponding generated videos.
• Score: 72.04671111142191
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Video generation models are increasingly capable of producing realistic videos, but they still struggle to generate videos that follow basic physical laws. Compounding this is a lack of reliable granular evaluation methods for localizing and specifying physical law violations in videos. We address this by introducing Physics Question Scene Graph (PQSG), a hierarchical question-based evaluation pipeline. PQSG evaluates generated videos by checking their faithfulness to a prompt across objects, actions, and adherence to physical laws using a graph-based hierarchy of questions generated by a vision-language model (VLM), guided by high-quality in-context examples. By representing questions as a graph, PQSG introduces logical dependencies within questions, ensuring that each query is contextually valid. Moreover, PQSG provides granular assessments of which qualities of the video violate physical plausibility constraints. We validate PQSG by creating FinePhyEval, a dataset with physics-based prompts and corresponding generated videos from diverse state-of-the-art video generation models (Sora 2, Veo 3, and Wan 2.1), with each video annotated across multiple categories by humans. Using FinePhyEval, we measure the correlation between PQSG's fine-grained scores and human judgments, showing higher overall correlations than prior work. We also find that PQSG ranks closed-source models higher than Wan 2.1 on physical realism. Lastly, we show that the annotations we provide in FinePhyEval can also be used for subtask evaluation: we benchmark two strong VLMs on generating and answering questions, finding that while models can create human-like questions, they still fall short of human performance in answering them.

Related papers

• Evidence-Backed Video Question Answering [90.93225758130426]
  Existing explainability efforts rely on textual rationales or sparse bounding boxes.<n>We propose Evidence-Backed Video Question Answering (E-VQA)<n>To support this, we introduce ST-Evidence, the first human-verified benchmark for both discriminative and generative pixel-level grounding.
  arXiv  Detail & Related papers  (2026-07-13T17:49:10Z)
• Physics-IQ Verified [58.94877959578469]
  We present a systematic audit of the Physics-IQ benchmark and expose shortcomings.<n>We propose three solutions that sharpen how we can measure physical understanding of video generative models.<n>Our resulting benchmark, Physics-IQ Verified, refines 57.6% of all samples and improves over 34.8% of prompts.
  arXiv  Detail & Related papers  (2026-06-17T11:23:52Z)
• PhyGround: Benchmarking Physical Reasoning in Generative World Models [46.53245929361594]
  We introduce PhyGround, a criteria-grounded benchmark for evaluating physical reasoning in video generation.<n>The benchmark contains 250 curated prompts, each augmented with an expected physical outcome, and a taxonomy of 13 physical laws.<n>We evaluate eight modern video generation models through a large-scale, quality-controlled human study.
  arXiv  Detail & Related papers  (2026-05-11T16:30:51Z)
• Can Your Model Separate Yolks with a Water Bottle? Benchmarking Physical Commonsense Understanding in Video Generation Models [14.187604603759784]
  We present PhysVidBench, a benchmark designed to evaluate the physical reasoning capabilities of text-to-video systems.<n>For each prompt, we generate videos using diverse state-of-the-art models and adopt a three-stage evaluation pipeline.<n> PhysVidBench provides a structured, interpretable framework for assessing physical commonsense in generative video models.
  arXiv  Detail & Related papers  (2025-07-21T17:30:46Z)
• ImplicitQA: Going beyond frames towards Implicit Video Reasoning [39.63171940350552]
  ImplicitQA is a novel benchmark designed to test VideoQA models on human-like implicit reasoning.<n>ImplicitQA comprises 1K meticulously annotated QA pairs drawn from 1K high-quality creative video clips.
  arXiv  Detail & Related papers  (2025-06-26T19:53:54Z)
• CausalVQA: A Physically Grounded Causal Reasoning Benchmark for Video Models [4.889577550694335]
  CausalVQA is a benchmark dataset for video question answering (VQA)<n>It consists of question-answer pairs that probe models' understanding of causality in the physical world.
  arXiv  Detail & Related papers  (2025-06-11T17:10:36Z)
• Towards Fine-Grained Video Question Answering [17.582244704442747]
  This paper introduces the Multi-Object Multi-Actor Question Answering (MOMA-QA) dataset.<n>With ground truth scene graphs and temporal interval annotations, MOMA-QA is ideal for developing models for fine-grained video understanding.<n>We present a novel video-language model, SGVLM, which incorporates a scene graph predictor, an efficient frame retriever, and a pre-trained large language model for temporal localization and fine-grained relationship understanding.
  arXiv  Detail & Related papers  (2025-03-10T01:02:01Z)
• VideoPhy-2: A Challenging Action-Centric Physical Commonsense Evaluation in Video Generation [66.58048825989239]
  VideoPhy-2 is an action-centric dataset for evaluating physical commonsense in generated videos.<n>We perform human evaluation that assesses semantic adherence, physical commonsense, and grounding of physical rules in the generated videos.<n>Our findings reveal major shortcomings, with even the best model achieving only 22% joint performance.
  arXiv  Detail & Related papers  (2025-03-09T22:49:12Z)
• TemporalBench: Benchmarking Fine-grained Temporal Understanding for Multimodal Video Models [75.42002690128486]
  TemporalBench is a new benchmark dedicated to evaluating fine-grained temporal understanding in videos. It consists of 10K video question-answer pairs, derived from 2K high-quality human annotations detailing the temporal dynamics in video clips. Results show that state-of-the-art models like GPT-4o achieve only 38.5% question answering accuracy on TemporalBench.
  arXiv  Detail & Related papers  (2024-10-14T17:59:58Z)
• Invariant Grounding for Video Question Answering [72.87173324555846]
  Video Question Answering (VideoQA) is the task of answering questions about a video. In leading VideoQA models, the typical learning objective, empirical risk minimization (ERM), latches on superficial correlations between video-question pairs and answers. We propose a new learning framework, Invariant Grounding for VideoQA (IGV), to ground the question-critical scene.
  arXiv  Detail & Related papers  (2022-06-06T04:37:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.