notesum.ai

Published at October 24

Robust Watermarking Using Generative Priors Against Image Editing: From Benchmarking to Advances

cs.CV

cs.AI

cs.CR

Released Date: October 24, 2024

Authors: Shilin Lu¹, Zihan Zhou¹, Jiayou Lu¹, Yuanzhi Zhu², Adams Wai-Kin Kong¹

Aff.: ¹Nanyang Technological University; ²ETH Zurich

Arxiv: https://arxiv.org/abs/2410.18775v1

Refer to caption

Config	Blurring	Watermark Encoder					PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$	FID $\downarrow$	TPR@0.1%FPR $\uparrow$ (%)
Config	Distortions	Backbone	Condition	Skip	Pretrained	Finetune	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$	FID $\downarrow$	Sto	Det	Pix2Pix	Ultra
Config A	\faTimes	Simple UNet	N.A.	N.A.	N.A.	\faTimes	38.21	0.9828	0.0148	1.69	54.61	66.86	64.24	32.62
Config B	\faCheck					\faTimes	35.85	0.9766	0.0257	2.12	86.85	92.28	80.98	62.14
Config C	\faCheck					\faCheck	31.24	0.9501	0.0458	4.67	98.59	99.29	96.01	84.60
Config D	\faCheck	SDXL-Turbo	ControlNet	\faTimes	\faCheck	\faTimes	32.68	0.9640	0.0298	2.87	90.82	94.89	91.86	70.69
Config E	\faCheck		Cond. Adaptor	\faTimes	\faCheck	\faTimes	36.76	0.9856	0.0102	0.53	90.86	94.78	92.88	70.68
Config F (VINE-B)	\faCheck		Cond. Adaptor	\faCheck	\faCheck	\faTimes	40.51	0.9954	0.0029	0.08	91.03	99.25	96.30	80.90
Config G (VINE-R)	\faCheck		Cond. Adaptor	\faCheck	\faCheck	\faCheck	37.34	0.9934	0.0063	0.15	99.66	99.98	97.46	86.86
Config H	\faCheck		Cond. Adaptor	\faCheck	\faTimes	\faCheck	35.18	0.9812	0.0137	1.03	99.67	99.92	96.13	84.66