Compositional Physical Reasoning of Objects and Events from Videos
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866918034478727168 |
|---|---|
| author | Chen, Zhenfang Dong, Shilong Yi, Kexin Li, Yunzhu Ding, Mingyu Torralba, Antonio Tenenbaum, Joshua B. Gan, Chuang |
| author_facet | Chen, Zhenfang Dong, Shilong Yi, Kexin Li, Yunzhu Ding, Mingyu Torralba, Antonio Tenenbaum, Joshua B. Gan, Chuang |
| contents | Understanding and reasoning about objects' physical properties in the natural world is a fundamental challenge in artificial intelligence. While some properties like colors and shapes can be directly observed, others, such as mass and electric charge, are hidden from the objects' visual appearance. This paper addresses the unique challenge of inferring these hidden physical properties from objects' motion and interactions and predicting corresponding dynamics based on the inferred physical properties. We first introduce the Compositional Physical Reasoning (ComPhy) dataset. For a given set of objects, ComPhy includes limited videos of them moving and interacting under different initial conditions. The model is evaluated based on its capability to unravel the compositional hidden properties, such as mass and charge, and use this knowledge to answer a set of questions. Besides the synthetic videos from simulators, we also collect a real-world dataset to show further test physical reasoning abilities of different models. We evaluate state-of-the-art video reasoning models on ComPhy and reveal their limited ability to capture these hidden properties, which leads to inferior performance. We also propose a novel neuro-symbolic framework, Physical Concept Reasoner (PCR), that learns and reasons about both visible and hidden physical properties from question answering. After training, PCR demonstrates remarkable capabilities. It can detect and associate objects across frames, ground visible and hidden physical properties, make future and counterfactual predictions, and utilize these extracted representations to answer challenging questions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2408_02687 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Compositional Physical Reasoning of Objects and Events from Videos Chen, Zhenfang Dong, Shilong Yi, Kexin Li, Yunzhu Ding, Mingyu Torralba, Antonio Tenenbaum, Joshua B. Gan, Chuang Computer Vision and Pattern Recognition Understanding and reasoning about objects' physical properties in the natural world is a fundamental challenge in artificial intelligence. While some properties like colors and shapes can be directly observed, others, such as mass and electric charge, are hidden from the objects' visual appearance. This paper addresses the unique challenge of inferring these hidden physical properties from objects' motion and interactions and predicting corresponding dynamics based on the inferred physical properties. We first introduce the Compositional Physical Reasoning (ComPhy) dataset. For a given set of objects, ComPhy includes limited videos of them moving and interacting under different initial conditions. The model is evaluated based on its capability to unravel the compositional hidden properties, such as mass and charge, and use this knowledge to answer a set of questions. Besides the synthetic videos from simulators, we also collect a real-world dataset to show further test physical reasoning abilities of different models. We evaluate state-of-the-art video reasoning models on ComPhy and reveal their limited ability to capture these hidden properties, which leads to inferior performance. We also propose a novel neuro-symbolic framework, Physical Concept Reasoner (PCR), that learns and reasons about both visible and hidden physical properties from question answering. After training, PCR demonstrates remarkable capabilities. It can detect and associate objects across frames, ground visible and hidden physical properties, make future and counterfactual predictions, and utilize these extracted representations to answer challenging questions. |
| title | Compositional Physical Reasoning of Objects and Events from Videos |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2408.02687 |