No abstract

No abstract

A drawback of most watermarking techniques is the need for some additional information in order to retrieve the watermark. Additionally, the robustness of the watermark decreases as the number of information bits stored in the image increases. We present a Watermarking technique which requires no information for decoding in addition to the watermarked image. The watermark is multi-level with few bits embedded robustly at low levels and longer watermark sequences embedded less robustly at higher levels. This allows detection of "tampered" and "attacked" images by detection of existence of the watermark at low levels and deterioration of the watermark at high levels. In order to prevent interference of the watermarks at different levels various image representation spaces are used. The watermark is shown to be non visible and robust.

Bob offers a face-detection web service where clients can submit their images for analysis. Alice would very much like to use the service, but is reluctant to reveal the content of her images to Bob. Bob, for his part, is reluctant to release his face detector, as he spent a lot of time, energy and money constructing it. Secure Multi-Party computations use cryptographic tools to solve this problem without leaking any information. Unfortunately, these methods are slow to compute and we introduce a couple of machine learning techniques that allow the parties to solve the problem while leaking a controlled amount of information. The first method is an information-bottleneck variant of AdaBoost that lets Bob find a subset of features that are enough for classifying an image patch, but not enough to actually reconstruct it. The second machine learning technique is active learning that allows Alice to construct an online classifier, based on a small number of calls to Bob's face detector. She can then use her online classifier as a fast rejector before using a cryptographically secure classifier on the remaining image patches.

No abstract