paper-with-me

Papers

CoPhy: Counterfactual Learning of Physical Dynamics

2019-09-26 · ICLR 2020 1 · Fabien Baradel, Natalia Neverova, Julien Mille, Greg Mori, Christian Wolf

Understanding causes and effects in mechanical systems is an essential component of reasoning in the physical world. This work poses a new problem of counterfactual learning of object mechanics from visual input. We develop the CoPhy benchmark to assess the capacity of the state-of-the-art models for causal physical reasoning in a synthetic 3D environment and propose a model for learning the physical dynamics in a counterfactual setting. Having observed a mechanical experiment that involves, for example, a falling tower of blocks, a set of bouncing balls or colliding objects, we learn to predict how its outcome is affected by an arbitrary intervention on its initial conditions, such as displacing one of the objects in the scene. The alternative future is predicted given the altered past and a latent representation of the confounders learned by the model in an end-to-end fashion with no supervision. We compare against feedforward video prediction baselines and show how observing alternative experiences allows the network to capture latent physical properties of the environment, which results in significantly more accurate predictions at the level of super human performance.

📄 PDF Abstract BibTeX arXiv:1909.12000

Code (1)

fabienbaradel/cophy pytorch

Tasks

counterfactualVideo Prediction

Similar Papers 제목 키워드 기반

Filtered-CoPhy: Unsupervised Learning of Counterfactual Physics in Pixel Space

2022-02-01 · ICLR 2022 4 · Steeven Janny, Fabien Baradel, Natalia Neverova, Madiha Nadri 외

Learning causal relationships in high-dimensional data (images, videos) is a hard task, as they are often defined on low dimensional manifolds and must be extracted from complex signals dominated by appearance, lighting,…

counterfactualCounterfactual ReasoningVideo Prediction

Causal Counterfactuals for Improving the Robustness of Reinforcement Learning

2022-11-02 · Tom He, Jasmina Gajcin, Ivana Dusparic

Reinforcement learning (RL) is used in various robotic applications. RL enables agents to learn tasks autonomously by interacting with the environment. The more critical the tasks are, the higher the demand for the robus…

Causal Inferencereinforcement-learningReinforcement LearningReinforcement Learning (RL)+1

Understanding Physical Dynamics with Counterfactual World Modeling

2023-12-11 · Rahul Venkatesh, Honglin Chen, Kevin Feigelis, Daniel M. Bear 외

The ability to understand physical dynamics is critical for agents to act in the world. Here, we use Counterfactual World Modeling (CWM) to extract vision structures for dynamics understanding. CWM uses a temporally-fact…

counterfactual

Distill to Think, Foresee to Act: Cognitive-Physical Reinforcement Learning for Autonomous Driving

2026-05-20 · Yang Wu, Qiang Meng, Zhaojiang Liu, Youquan Liu 외 arxiv

Current end-to-end autonomous driving models are fundamentally constrained by the behavioral cloning ceiling of imitation learning. While reinforcement learning offers a path to smarter autonomy, it demands two missing p…

Reinforcement LearningAutonomous Driving

PhyLatent: Learning Dynamics-Relevant Representations for JEPA World Models

2026-08-06 · Xi Zeng, Haojie Ren, Ziying Song arxiv

We propose PhyLatent, a dynamics-relevant training objective for JointEmbedding Predictive Architecture (JEPA) world models. Our key observation is that preventing global latent collapse does not ensure that a representa…