FiLM-CAM secures image watermarks with secret key based control
FiLM-CAM: Keyed Feature Modulation for Conditional-Access Watermarking
Computer Vision and Pattern Recognition
Summary
Protecting images with hidden watermarks can be tricky, especially when only authorized people should add, read, or remove them. The authors created a system called FiLM-CAM that uses a secret key to change how watermarks are hidden inside images at a deep feature level, not just in the pixels. This means only someone with the right key can find or erase the watermark properly. They also trained the system so it works well when using the correct key, but fails if the wrong key is used. This method works across different kinds of watermarking systems, making it flexible and reliable.
What this means in practice
- •For digital content protection teams: Enable controlled access to image watermarks ensuring only authorized parties can embed, extract, or remove the watermark using secret keys.
- •For forensic analysts: Identify the presence of protected watermarks in images while preventing unauthorized decoding or erasure, aiding copyright enforcement.
Authors
Yash Kulthe, Vishal Asnani, Shruti Agarwal, Andrew Gilbert, John Collomosse
Abstract
We propose FiLM-CAM; a conditional-access image watermarking approach that enforces authorized encoding, decoding, and removal through keyed feature modulation. A secret key conditions a Feature-wise Linear Modulation (FiLM) mechanism that transforms the internal representation of the watermark signal before embedding, thereby coupling the payload to the key at the feature, rather than pixel level. The same key must be presented to condition the decoder to correctly recover the payload embedded in an image. To enforce conditional access, we train the system to maximize the performance gap between authorized and unauthorized use, ensuring reliable extraction under the correct key while inducing failure under incorrect keys. In addition, we design our encoder to estimate the watermark residual invariant to the presence of pre-existing watermarks. This design enables conditional watermark removal via simple re-encoding and subtraction of the residual using iterative refinement, contingent on knowledge of the secret key. We evaluate FiLM-CAM across multiple watermarking encoder-decoder backbones, demonstrating that keyed feature modulation serves as an effective, architecture-agnostic conditional access module (CAM) for image watermarking.