Divide and Conquer: Reliable Multi-View Evidential Learning for Deepfake Detection

Xiaolu Kang, Zhongyuan Wang, Jikang Cheng, Baojin Huang, Zhanhe Lei, Gang Wu, Qin Zou, Qian Wang
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:55723-55736, 2026.

Abstract

With the evolution of generative models, deepfakes have achieved near-perfect semantic realism, leaving forensic traces only in subtle structural anomalies. However, existing single-view paradigms often fail to generalize, as dominant semantic features overwhelm subtle artifact cues within entangled representations. This imbalance leads to overconfident yet brittle predictions—a phenomenon we term the Semantic Masking Effect. To address this challenge, we propose a reliable framework called Divide-and-Conquer Multi-View Evidential Learning (DiCoME) for Deepfake Detection. In the "Divide” phase, we employ Geometric View Purification to decompose the entangled representation space through principled geometric projection. This process suppresses semantic interference within artifact-sensitive representations, forming the foundation for decorrelated yet complementary semantic and artifact views. In the "Conquer” phase, we leverage Uncertainty-Aware Evidential Learning to synthesize these distinct views. By explicitly modeling the "epistemic conflict” between semantic and artifact cues, this mechanism provides calibrated uncertainty estimates instead of forcing rigid deterministic decisions. Extensive experiments across multiple benchmarks demonstrate that our method consistently outperforms existing approaches in generalization performance, while providing reliable uncertainty estimation for trustworthy deepfake detection. Code is available at https://github.com/kxl0825/DiCoME.git.

Cite this Paper


BibTeX
@InProceedings{pmlr-v306-kang26o, title = {Divide and Conquer: Reliable Multi-View Evidential Learning for Deepfake Detection}, author = {Kang, Xiaolu and Wang, Zhongyuan and Cheng, Jikang and Huang, Baojin and Lei, Zhanhe and Wu, Gang and Zou, Qin and Wang, Qian}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {55723--55736}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/kang26o/kang26o.pdf}, url = {https://proceedings.mlr.press/v306/kang26o.html}, abstract = {With the evolution of generative models, deepfakes have achieved near-perfect semantic realism, leaving forensic traces only in subtle structural anomalies. However, existing single-view paradigms often fail to generalize, as dominant semantic features overwhelm subtle artifact cues within entangled representations. This imbalance leads to overconfident yet brittle predictions—a phenomenon we term the Semantic Masking Effect. To address this challenge, we propose a reliable framework called Divide-and-Conquer Multi-View Evidential Learning (DiCoME) for Deepfake Detection. In the "Divide” phase, we employ Geometric View Purification to decompose the entangled representation space through principled geometric projection. This process suppresses semantic interference within artifact-sensitive representations, forming the foundation for decorrelated yet complementary semantic and artifact views. In the "Conquer” phase, we leverage Uncertainty-Aware Evidential Learning to synthesize these distinct views. By explicitly modeling the "epistemic conflict” between semantic and artifact cues, this mechanism provides calibrated uncertainty estimates instead of forcing rigid deterministic decisions. Extensive experiments across multiple benchmarks demonstrate that our method consistently outperforms existing approaches in generalization performance, while providing reliable uncertainty estimation for trustworthy deepfake detection. Code is available at https://github.com/kxl0825/DiCoME.git.} }
Endnote
%0 Conference Paper %T Divide and Conquer: Reliable Multi-View Evidential Learning for Deepfake Detection %A Xiaolu Kang %A Zhongyuan Wang %A Jikang Cheng %A Baojin Huang %A Zhanhe Lei %A Gang Wu %A Qin Zou %A Qian Wang %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-kang26o %I PMLR %P 55723--55736 %U https://proceedings.mlr.press/v306/kang26o.html %V 306 %X With the evolution of generative models, deepfakes have achieved near-perfect semantic realism, leaving forensic traces only in subtle structural anomalies. However, existing single-view paradigms often fail to generalize, as dominant semantic features overwhelm subtle artifact cues within entangled representations. This imbalance leads to overconfident yet brittle predictions—a phenomenon we term the Semantic Masking Effect. To address this challenge, we propose a reliable framework called Divide-and-Conquer Multi-View Evidential Learning (DiCoME) for Deepfake Detection. In the "Divide” phase, we employ Geometric View Purification to decompose the entangled representation space through principled geometric projection. This process suppresses semantic interference within artifact-sensitive representations, forming the foundation for decorrelated yet complementary semantic and artifact views. In the "Conquer” phase, we leverage Uncertainty-Aware Evidential Learning to synthesize these distinct views. By explicitly modeling the "epistemic conflict” between semantic and artifact cues, this mechanism provides calibrated uncertainty estimates instead of forcing rigid deterministic decisions. Extensive experiments across multiple benchmarks demonstrate that our method consistently outperforms existing approaches in generalization performance, while providing reliable uncertainty estimation for trustworthy deepfake detection. Code is available at https://github.com/kxl0825/DiCoME.git.
APA
Kang, X., Wang, Z., Cheng, J., Huang, B., Lei, Z., Wu, G., Zou, Q. & Wang, Q.. (2026). Divide and Conquer: Reliable Multi-View Evidential Learning for Deepfake Detection. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:55723-55736 Available from https://proceedings.mlr.press/v306/kang26o.html.

Related Material