# Compositional 4D Dynamic Scenes Understanding with Physics Priors for Video Question Answering

arXiv: https://arxiv.org/abs/2406.00622 · HTML: https://mlsanity.com/paper/2406.00622/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- Placed in Embodied and 3D vision-language models, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 27 of its 45 references that are in the map.
- First posted on arXiv 2 Jun 2024.

## Placed in

- Multimodal LLMs / vision-language models / [Embodied and 3D vision-language models](https://mlsanity.com/mllm/embodied-and-3d-vision-language-models/)

## Builds on (most cited first, 27 of 27)

- [Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks](https://mlsanity.com/paper/1506.01497/) (2015)
- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://mlsanity.com/paper/2210.08402/) (2022)
- [FiLM: Visual Reasoning with a General Conditioning Layer](https://mlsanity.com/paper/1709.07871/) (2017)
- [Video-LLaVA: Learning United Visual Representation by Alignment Before Projection](https://mlsanity.com/paper/2311.10122/) (2023)
- [Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval](https://mlsanity.com/paper/2104.00650/) (2021)
- [The "something something" video database for learning and evaluating visual common sense](https://mlsanity.com/paper/1706.04261/) (2017)
- [CLEVRER: CoLlision Events for Video REpresentation and Reasoning](https://mlsanity.com/paper/1910.01442/) (2019)
- [InternVideo: General Video Foundation Models via Generative and Discriminative Learning](https://mlsanity.com/paper/2212.03191/) (2022)
- [Soft Rasterizer: A Differentiable Renderer for Image-based 3D Reasoning](https://mlsanity.com/paper/1904.01786/) (2019)
- [PLLaVA : Parameter-free LLaVA Extension from Images to Videos for Video Dense Captioning](https://mlsanity.com/paper/2404.16994/) (2024)
- [STAR: A Benchmark for Situated Reasoning in Real-World Videos](https://mlsanity.com/paper/2405.09711/) (2024)
- [Physion: Evaluating Physical Prediction from Vision in Humans and Machines](https://mlsanity.com/paper/2106.08261/) (2021)
- [CATER: A diagnostic dataset for Compositional Actions and TEmporal Reasoning](https://mlsanity.com/paper/1910.04744/) (2019)
- [Super-CLEVR: A Virtual Benchmark to Diagnose Domain Robustness in Visual Reasoning](https://mlsanity.com/paper/2212.00259/) (2022)
- [ComPhy: Compositional Physical Reasoning of Objects and Events from Videos](https://mlsanity.com/paper/2205.01089/) (2022)
- [Dynamic Visual Reasoning by Learning Differentiable Physics Models from Video and Language](https://mlsanity.com/paper/2110.15358/) (2021)
- [Physion++: Evaluating Physical Scene Understanding that Requires Online Inference of Different Physical Properties](https://mlsanity.com/paper/2306.15668/) (2023)
- [CRAFT: A Benchmark for Causal Reasoning About Forces and inTeractions](https://mlsanity.com/paper/2012.04293/) (2020)
- [NeMo: Neural Mesh Models of Contrastive Features for Robust 3D Pose Estimation](https://mlsanity.com/paper/2101.12378/) (2021)
- [Learning to Answer Visual Questions from Web Videos](https://mlsanity.com/paper/2205.05019/) (2022)
- [ContPhy: Continuum Physical Concept Learning and Reasoning from Videos](https://mlsanity.com/paper/2402.06119/) (2024)
- [Robust Category-Level 6D Pose Estimation with Coarse-to-Fine Rendering of Neural Features](https://mlsanity.com/paper/2209.05624/) (2022)
- [3D-Aware Visual Question Answering about Parts, Poses and Occlusions](https://mlsanity.com/paper/2310.17914/) (2023)
- [CRIPP-VQA: Counterfactual Reasoning about Implicit Physical Properties via Video Question Answering](https://mlsanity.com/paper/2211.03779/) (2022)
- [CoKe: Localized Contrastive Learning for Robust Keypoint Detection](https://mlsanity.com/paper/2009.14115/) (2020)
- [NOVUM: Neural Object Volumes for Robust Object Classification](https://mlsanity.com/paper/2305.14668/) (2023)
- [Source-Free and Image-Only Unsupervised Domain Adaptation for Category Level Object Pose Estimation](https://mlsanity.com/paper/2401.10848/) (2024)

## Built on by (most cited first, 18 of 18)

- [3DSRBench: A Comprehensive 3D Spatial Reasoning Benchmark](https://mlsanity.com/paper/2412.07825/) (2024)
- [SpatialReasoner: Towards Explicit and Generalizable 3D Spatial Reasoning](https://mlsanity.com/paper/2504.20024/) (2025)
- [IntPhys 2: Benchmarking Intuitive Physics Understanding In Complex Synthetic Environments](https://mlsanity.com/paper/2506.09849/) (2025)
- [VLM4D: Towards Spatiotemporal Awareness in Vision Language Models](https://mlsanity.com/paper/2508.02095/) (2025)
- [How Far are VLMs from Visual Spatial Intelligence? A Benchmark-Driven Perspective](https://mlsanity.com/paper/2509.18905/) (2025)
- [SpatialLLM: A Compound 3D-Informed Design towards Spatially-Intelligent Large Multimodal Models](https://mlsanity.com/paper/2505.00788/) (2025)
- [SpatialThinker: Reinforcing Scene Graph-Grounded Spatial Reasoning via Dense Rewards](https://mlsanity.com/paper/2511.07403/) (2025)
- [Grounding Creativity in Physics: A Brief Survey of Physical Priors in AIGC](https://mlsanity.com/paper/2502.07007/) (2025)
- [Thinking in Dynamics: How Multimodal Large Language Models Perceive, Track, and Reason Dynamics in Physical 4D World](https://mlsanity.com/paper/2603.12746/) (2026)
- [PhysInOne: Visual Physics Learning and Reasoning in One Suite](https://mlsanity.com/paper/2604.09415/) (2026)
- [PhyBlock: A Progressive Benchmark for Physical Understanding and Planning via 3D Block Assembly](https://mlsanity.com/paper/2506.08708/) (2025)
- [Seeing the Wind from a Falling Leaf](https://mlsanity.com/paper/2512.00762/) (2025)
- [LMM-Track4D: Eliciting 4D Dynamic Reasoning in LMMs via Trajectory-Grounded Dialogue](https://mlsanity.com/paper/2605.19390/) (2026)
- [Augmented Vision-Language Models: A Systematic Review](https://mlsanity.com/paper/2507.22933/) (2025)
- [Bridging the Gap Between Multimodal Foundation Models and World Models](https://mlsanity.com/paper/2510.03727/) (2025)
- [Order from Chaos: Physical World Understanding from Glitchy Gameplay Videos](https://mlsanity.com/paper/2601.16471/) (2026)
- [IMBench: A Benchmark for Intuitive Robotic Manipulation](https://mlsanity.com/paper/2607.15641/) (2026)
- [VGenST-Bench: A Benchmark for Spatio-Temporal Reasoning via Active Video Synthesis](https://mlsanity.com/paper/2605.22570/) (2026)

## Abstract (arXiv)

For vision-language models (VLMs), understanding the dynamic properties of objects and their interactions in 3D scenes from videos is crucial for effective reasoning about high-level temporal and action semantics. Although humans are adept at understanding these properties by constructing 3D and temporal (4D) representations of the world, current video understanding models struggle to extract these dynamic semantics, arguably because these models use cross-frame reasoning without underlying knowledge of the 3D/4D scenes. In this work, we introduce DynSuperCLEVR, the first video question answering dataset that focuses on language understanding of the dynamic properties of 3D objects. We concentrate on three physical concepts -- velocity, acceleration, and collisions within 4D scenes. We further generate three types of questions, including factual queries, future predictions, and counterfactual reasoning that involve different aspects of reasoning about these 4D dynamic properties. To further demonstrate the importance of explicit scene representations in answering these 4D dynamics questions, we propose NS-4DPhysics, a Neural-Symbolic VideoQA model integrating Physics prior for 4D dynamic properties with explicit scene representation of videos. Instead of answering the questions directly from the video text input, our method first estimates the 4D world states with a 3D generative model powered by physical priors, and then uses neural symbolic reasoning to answer the questions based on the 4D world states. Our evaluation on all three types of questions in DynSuperCLEVR shows that previous video question answering models and large multimodal models struggle with questions about 4D dynamics, while our NS-4DPhysics significantly outperforms previous state-of-the-art models. Our code and data are released in https://xingruiwang.github.io/projects/DynSuperCLEVR/.