FUSE: Full-spectrum Unlearnable Examples via Spectral Equalization

Jiale Cai, Gezheng Xu, Zhihao Li, Ruiyi Fang, Ruizhi Pu, Di Wu, Qicheng Lao, Charles Ling, Boyu Wang
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:10764-10788, 2026.

Abstract

Unlearnable examples (UEs) protect training data by injecting imperceptible perturbations so that models fail to extract exploitable representations. In this paper, we reveal that existing UEs exhibit a critical failure once low-pass filtering is applied, indicating that the effective perturbation signals for unlearnability concentrate predominantly in high frequencies. Hence, we argue that reliable UEs should remain effective across the full spectrum. To this end, we propose Full-spectrum Unlearnable Examples via Spectral Equalization (FUSE), which aims to generate spectrum-agnostic perturbations by equalizing the contributions from different bands and enforcing cross-band consistency. Specifically, FUSE adopts a Random Spectral Masking (RSM) strategy during generator training, which randomly removes a contiguous frequency band, forcing the remaining bands to maintain unlearnability. In addition, FUSE further integrates Cross-Band Guidance (CBG), which enforces mutual consistency between high- and low-frequency components, thereby further enhancing low-frequency unlearnability and regulating high-frequency perturbations to preserve the semantic fidelity of images. Extensive experiments across multiple datasets, architectures, and spectral filtering demonstrate the strong protection achieved by FUSE.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-cai26m, title = {{FUSE}: Full-spectrum Unlearnable Examples via Spectral Equalization}, author = {Cai, Jiale and Xu, Gezheng and Li, Zhihao and Fang, Ruiyi and Pu, Ruizhi and Wu, Di and Lao, Qicheng and Ling, Charles and Wang, Boyu}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {10764--10788}, 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/cai26m/cai26m.pdf}, url = {https://proceedings.mlr.press/v306/cai26m.html}, abstract = {Unlearnable examples (UEs) protect training data by injecting imperceptible perturbations so that models fail to extract exploitable representations. In this paper, we reveal that existing UEs exhibit a critical failure once low-pass filtering is applied, indicating that the effective perturbation signals for unlearnability concentrate predominantly in high frequencies. Hence, we argue that reliable UEs should remain effective across the full spectrum. To this end, we propose Full-spectrum Unlearnable Examples via Spectral Equalization (FUSE), which aims to generate spectrum-agnostic perturbations by equalizing the contributions from different bands and enforcing cross-band consistency. Specifically, FUSE adopts a Random Spectral Masking (RSM) strategy during generator training, which randomly removes a contiguous frequency band, forcing the remaining bands to maintain unlearnability. In addition, FUSE further integrates Cross-Band Guidance (CBG), which enforces mutual consistency between high- and low-frequency components, thereby further enhancing low-frequency unlearnability and regulating high-frequency perturbations to preserve the semantic fidelity of images. Extensive experiments across multiple datasets, architectures, and spectral filtering demonstrate the strong protection achieved by FUSE.} }
Endnote
%0 Conference Paper %T FUSE: Full-spectrum Unlearnable Examples via Spectral Equalization %A Jiale Cai %A Gezheng Xu %A Zhihao Li %A Ruiyi Fang %A Ruizhi Pu %A Di Wu %A Qicheng Lao %A Charles Ling %A Boyu 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-cai26m %I PMLR %P 10764--10788 %U https://proceedings.mlr.press/v306/cai26m.html %V 306 %X Unlearnable examples (UEs) protect training data by injecting imperceptible perturbations so that models fail to extract exploitable representations. In this paper, we reveal that existing UEs exhibit a critical failure once low-pass filtering is applied, indicating that the effective perturbation signals for unlearnability concentrate predominantly in high frequencies. Hence, we argue that reliable UEs should remain effective across the full spectrum. To this end, we propose Full-spectrum Unlearnable Examples via Spectral Equalization (FUSE), which aims to generate spectrum-agnostic perturbations by equalizing the contributions from different bands and enforcing cross-band consistency. Specifically, FUSE adopts a Random Spectral Masking (RSM) strategy during generator training, which randomly removes a contiguous frequency band, forcing the remaining bands to maintain unlearnability. In addition, FUSE further integrates Cross-Band Guidance (CBG), which enforces mutual consistency between high- and low-frequency components, thereby further enhancing low-frequency unlearnability and regulating high-frequency perturbations to preserve the semantic fidelity of images. Extensive experiments across multiple datasets, architectures, and spectral filtering demonstrate the strong protection achieved by FUSE.
APA
Cai, J., Xu, G., Li, Z., Fang, R., Pu, R., Wu, D., Lao, Q., Ling, C. & Wang, B.. (2026). FUSE: Full-spectrum Unlearnable Examples via Spectral Equalization. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:10764-10788 Available from https://proceedings.mlr.press/v306/cai26m.html.

Related Material