arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2302.00869cs.LGcs.CVcs.DMstat.ME
收藏

通过因果干预实现潜在表征的解耦

Disentanglement of Latent Representations via Causal Interventions

University of Auckland(奥克兰大学)

Gaël Gendron, Michael Witbrock, Gillian Dobbie

AI总结:

本文提出一种结合因果理论与向量量化变分自编码器的解耦方法,通过因果图干预实现图像单一因素的原子转换,并引入动作检索任务,在不平衡数据下也能精确操控高层语义属性。

摘要:

生成图像等数据的过程由独立且未知的变化因素控制。对这些变量的提取已在解耦、因果表示学习和独立成分分析领域得到广泛研究。近年来,将这些领域融合在一起的方法取得了巨大成功。解耦问题可以不直接表示变化因素,而是被视为寻找对一幅图像施加的、能够引起单一因素变化的干预。基于这一假设,我们引入了一种受因果动力学启发的解耦新方法,将因果理论与向量量化变分自编码器相结合。我们的模型将量化向量视为因果变量,并将它们连接在一个因果图中。该模型对因果图执行因果干预,并生成影响图像中唯一变化因素的原子转换。我们还引入了一项新任务——动作检索,即找出导致两幅图像之间转换的动作。我们在标准的合成和真实世界解耦数据集上测试了我们的方法。我们表明,即使在数据分布不平衡的情况下,该方法也能有效解耦变化因素,并对图像的高层语义属性执行精确干预,同时不影响图像质量。

英文摘要:

The process of generating data such as images is controlled by independent and unknown factors of variation. The retrieval of these variables has been studied extensively in the disentanglement, causal representation learning, and independent component analysis fields. Recently, approaches merging these domains together have shown great success. Instead of directly representing the factors of variation, the problem of disentanglement can be seen as finding the interventions on one image that yield a change to a single factor. Following this assumption, we introduce a new method for disentanglement inspired by causal dynamics that combines causality theory with vector-quantized variational autoencoders. Our model considers the quantized vectors as causal variables and links them in a causal graph. It performs causal interventions on the graph and generates atomic transitions affecting a unique factor of variation in the image. We also introduce a new task of action retrieval that consists of finding the action responsible for the transition between two images. We test our method on standard synthetic and real-world disentanglement datasets. We show that it can effectively disentangle the factors of variation and perform precise interventions on high-level semantic attributes of an image without affecting its quality, even with imbalanced data distributions.

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。

↑