Enhancing Generalization of Invisible Facial Privacy Cloak via Gradient Accumulation
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916080356687872 |
|---|---|
| author | Liu, Xuannan Zhong, Yaoyao Deng, Weihong Shi, Hongzhi Cui, Xingchen Yin, Yunfeng Wen, Dongchao |
| author_facet | Liu, Xuannan Zhong, Yaoyao Deng, Weihong Shi, Hongzhi Cui, Xingchen Yin, Yunfeng Wen, Dongchao |
| contents | The blooming of social media and face recognition (FR) systems has increased people's concern about privacy and security. A new type of adversarial privacy cloak (class-universal) can be applied to all the images of regular users, to prevent malicious FR systems from acquiring their identity information. In this work, we discover the optimization dilemma in the existing methods -- the local optima problem in large-batch optimization and the gradient information elimination problem in small-batch optimization. To solve these problems, we propose Gradient Accumulation (GA) to aggregate multiple small-batch gradients into a one-step iterative gradient to enhance the gradient stability and reduce the usage of quantization operations. Experiments show that our proposed method achieves high performance on the Privacy-Commons dataset against black-box face recognition models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_01575 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Enhancing Generalization of Invisible Facial Privacy Cloak via Gradient Accumulation Liu, Xuannan Zhong, Yaoyao Deng, Weihong Shi, Hongzhi Cui, Xingchen Yin, Yunfeng Wen, Dongchao Computer Vision and Pattern Recognition The blooming of social media and face recognition (FR) systems has increased people's concern about privacy and security. A new type of adversarial privacy cloak (class-universal) can be applied to all the images of regular users, to prevent malicious FR systems from acquiring their identity information. In this work, we discover the optimization dilemma in the existing methods -- the local optima problem in large-batch optimization and the gradient information elimination problem in small-batch optimization. To solve these problems, we propose Gradient Accumulation (GA) to aggregate multiple small-batch gradients into a one-step iterative gradient to enhance the gradient stability and reduce the usage of quantization operations. Experiments show that our proposed method achieves high performance on the Privacy-Commons dataset against black-box face recognition models. |
| title | Enhancing Generalization of Invisible Facial Privacy Cloak via Gradient Accumulation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2401.01575 |