arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2510.14049cs.LGcs.MS
收藏

CausalVerse:基于可配置高保真模拟的因果表示学习基准

CausalVerse: Benchmarking Causal Representation Learning with Configurable High-Fidelity Simulations

Carnegie Mellon University(卡内基梅隆大学);Mohamed bin Zayed University of Artificial Intelligence(穆罕默德·本·扎耶德人工智能大学)

Guangyi Chen, Yunlong Deng, Peiyuan Zhu, Yan Li, Yifan Shen, Zijian Li, Kun Zhang

AI总结:

本文提出CausalVerse基准,利用高保真模拟视觉数据(含20万图像和300万帧)提供可配置的因果结构真值,以解决CRL评估中真实性与精度难以兼顾的问题,并评估了代表性方法。

摘要:

因果表示学习(CRL)旨在揭示数据生成过程,并识别潜在的因果变量及其关系,其评估由于需要已知的因果变量和因果结构真值而固有地具有挑战性。现有评估通常依赖于简单的合成数据集或现实世界任务的下游性能,普遍面临真实性与评估精度之间的两难困境。在本文中,我们引入了一个新的CRL基准,使用高保真模拟视觉数据,既保留了现实的视觉复杂性,更重要的是,提供了对真实因果生成过程的访问。该数据集包含约20万张图像和300万帧视频,涵盖四个领域的24个子场景:静态图像生成、动态物理模拟、机器人操作和交通情况分析。这些场景从静态到动态设置、从简单到复杂结构、从单一到多智能体交互,提供了一个全面的测试平台,有望弥合严格评估与现实世界适用性之间的差距。此外,我们提供对底层因果结构的灵活访问,允许用户修改或配置它们以符合CRL中的所需假设,例如可用的领域标签、时间依赖性或干预历史。利用该基准,我们评估了跨不同范式的代表性CRL方法,并提供了实证见解,以帮助从业者和新手选择或扩展适当的CRL框架,从而妥善处理可以从CRL视角受益的特定类型的现实问题。欢迎访问我们的:项目页面:https://causal-verse.github.io/,数据集:https://huggingface.co/CausalVerse。

英文摘要:

Causal Representation Learning (CRL) aims to uncover the data-generating process and identify the underlying causal variables and relations, whose evaluation remains inherently challenging due to the requirement of known ground-truth causal variables and causal structure. Existing evaluations often rely on either simplistic synthetic datasets or downstream performance on real-world tasks, generally suffering a dilemma between realism and evaluative precision. In this paper, we introduce a new benchmark for CRL using high-fidelity simulated visual data that retains both realistic visual complexity and, more importantly, access to ground-truth causal generating processes. The dataset comprises around 200 thousand images and 3 million video frames across 24 sub-scenes in four domains: static image generation, dynamic physical simulations, robotic manipulations, and traffic situation analysis. These scenarios range from static to dynamic settings, simple to complex structures, and single to multi-agent interactions, offering a comprehensive testbed that hopefully bridges the gap between rigorous evaluation and real-world applicability. In addition, we provide flexible access to the underlying causal structures, allowing users to modify or configure them to align with the required assumptions in CRL, such as available domain labels, temporal dependencies, or intervention histories. Leveraging this benchmark, we evaluated representative CRL methods across diverse paradigms and offered empirical insights to assist practitioners and newcomers in choosing or extending appropriate CRL frameworks to properly address specific types of real problems that can benefit from the CRL perspective. Welcome to visit our: Project page:https://causal-verse.github.io/, Dataset:https://huggingface.co/CausalVerse.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑