Diffusion Counterfactual Generation with Semantic Abduction

Rajat R Rasal, Avinash Kori, Fabio De Sousa Ribeiro, Tian Xia, Ben Glocker
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:51201-51228, 2025.

Abstract

Counterfactual image generation presents significant challenges, including preserving identity, maintaining perceptual quality, and ensuring faithfulness to an underlying causal model. While existing auto-encoding frameworks admit semantic latent spaces which can be manipulated for causal control, they struggle with scalability and fidelity. Advancements in diffusion models present opportunities for improving counterfactual image editing, having demonstrated state-of-the-art visual quality, human-aligned perception and representation learning capabilities. Here, we present a suite of diffusion-based causal mechanisms, introducing the notions of spatial, semantic and dynamic abduction. We propose a general framework that integrates semantic representations into diffusion models through the lens of Pearlian causality to edit images via a counterfactual reasoning process. To the best of our knowledge, ours is the first work to consider high-level semantic identity preservation for diffusion counterfactuals and to demonstrate how semantic control enables principled trade-offs between faithful causal control and identity preservation.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-rasal25a, title = {Diffusion Counterfactual Generation with Semantic Abduction}, author = {Rasal, Rajat R and Kori, Avinash and De Sousa Ribeiro, Fabio and Xia, Tian and Glocker, Ben}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {51201--51228}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/rasal25a/rasal25a.pdf}, url = {https://proceedings.mlr.press/v267/rasal25a.html}, abstract = {Counterfactual image generation presents significant challenges, including preserving identity, maintaining perceptual quality, and ensuring faithfulness to an underlying causal model. While existing auto-encoding frameworks admit semantic latent spaces which can be manipulated for causal control, they struggle with scalability and fidelity. Advancements in diffusion models present opportunities for improving counterfactual image editing, having demonstrated state-of-the-art visual quality, human-aligned perception and representation learning capabilities. Here, we present a suite of diffusion-based causal mechanisms, introducing the notions of spatial, semantic and dynamic abduction. We propose a general framework that integrates semantic representations into diffusion models through the lens of Pearlian causality to edit images via a counterfactual reasoning process. To the best of our knowledge, ours is the first work to consider high-level semantic identity preservation for diffusion counterfactuals and to demonstrate how semantic control enables principled trade-offs between faithful causal control and identity preservation.} }
Endnote
%0 Conference Paper %T Diffusion Counterfactual Generation with Semantic Abduction %A Rajat R Rasal %A Avinash Kori %A Fabio De Sousa Ribeiro %A Tian Xia %A Ben Glocker %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-rasal25a %I PMLR %P 51201--51228 %U https://proceedings.mlr.press/v267/rasal25a.html %V 267 %X Counterfactual image generation presents significant challenges, including preserving identity, maintaining perceptual quality, and ensuring faithfulness to an underlying causal model. While existing auto-encoding frameworks admit semantic latent spaces which can be manipulated for causal control, they struggle with scalability and fidelity. Advancements in diffusion models present opportunities for improving counterfactual image editing, having demonstrated state-of-the-art visual quality, human-aligned perception and representation learning capabilities. Here, we present a suite of diffusion-based causal mechanisms, introducing the notions of spatial, semantic and dynamic abduction. We propose a general framework that integrates semantic representations into diffusion models through the lens of Pearlian causality to edit images via a counterfactual reasoning process. To the best of our knowledge, ours is the first work to consider high-level semantic identity preservation for diffusion counterfactuals and to demonstrate how semantic control enables principled trade-offs between faithful causal control and identity preservation.
APA
Rasal, R.R., Kori, A., De Sousa Ribeiro, F., Xia, T. & Glocker, B.. (2025). Diffusion Counterfactual Generation with Semantic Abduction. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:51201-51228 Available from https://proceedings.mlr.press/v267/rasal25a.html.

Related Material