论文信息 - Protecting Visual Secrets Using Adversarial Nets

Protecting Visual Secrets Using Adversarial Nets

Protecting visual secrets is an important problem due to the prevalence of cameras that continuously monitor our surroundings. Any viable solution to this problem should also minimize the impact on the utility of applications that use images. In this work, we build on the existing work of adversarial learning to design a perturbation mechanism that jointly optimizes privacy and utility objectives. We provide a feasibility study of the proposed mechanism and present ideas on developing a privacy framework based on the adversarial perturbation mechanism.

Ashwin Machanavajjhala | Landon P. Cox | Nisarg Raval | Ashwin Machanavajjhala | Nisarg Raval

[1] Helen J. Wang,et al. Enabling Fine-Grained Permissions for Augmented Reality Applications with Recognizers , 2013, USENIX Security Symposium.

[2] Seong Joon Oh,et al. Faceless Person Recognition: Privacy Implications in Social Media , 2016, ECCV.

[3] Soumith Chintala,et al. Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks , 2015, ICLR.

[4] Vitaly Shmatikov,et al. Defeating Image Obfuscation with Deep Learning , 2016, ArXiv.

[5] Martín Abadi,et al. Learning to Protect Communications with Adversarial Neural Cryptography , 2016, ArXiv.

[6] Yoshua Bengio,et al. Generative Adversarial Nets , 2014, NIPS.

[7] Oriol Vinyals,et al. Matching Networks for One Shot Learning , 2016, NIPS.

[8] Noboru Murata,et al. Neural Network with Unbounded Activation Functions is Universal Approximator , 2015, 1505.03654.

[9] David J. Crandall,et al. PlaceAvoider: Steering First-Person Cameras away from Sensitive Spaces , 2014, NDSS.

[10] Amos J. Storkey,et al. Censoring Representations with an Adversary , 2015, ICLR.

[11] Alex Krizhevsky,et al. Learning Multiple Layers of Features from Tiny Images , 2009 .