Causal3D: A Comprehensive Benchmark for Causal Learning from Visual Data

Causal3D is a benchmark for evaluating causal representation learning and reasoning from visual data.

Recommended citation: Disheng Liu, Yiran Qiao, Wuche Liu, Yiren Lu, Yunlai Zhou, Tuo Liang, Yu Yin, and Jing Ma. "Causal3D: A Comprehensive Benchmark for Causal Learning from Visual Data." arXiv:2503.04852, 2025.
Download Paper