Our recently introduced JPEG steganographic method called Yet Another Steganographic Scheme (YASS) can resist blind steganalysis by embedding data in the discrete cosine transform (DCT) domain in randomly chosen image blocks. To maximize the embedding rate for a given image and a specified attack channel, the redundancy factor used by the repeat-accumulate (RA) code based error correction frame...

This paper aims to deal with the practical shortages of nearest neighbor classifier. We define a quantitative criterion of embedding quality assessment for nearest neighbor classification, and present a method called NNMap to construct a good embedding. Furthermore, an efficient distance is obtained in the embedded vector space, which could speed up nearest neighbor classification. The quantita...

In this work, we report a new quantum mechanical:quantum mechanical (QM:QM) method which provides explicit electronic polarization of the high-level region by using the Löwdin atomic charges from the low-level region. This provides an embedding potential which naturally evolves with changes in nuclear geometry. However, this coupling of the high-level and low-level regions introduces complicati...

Estimation of the Embedding capacity is an important problem specifically in reversible multi-pass watermarking and is required for analysis before any image can be watermarked. In this paper, we propose an efficient method for estimating the embedding capacity of a given cover image under multi-pass embedding, without actually embedding the watermark. We demonstrate this for a class of reversi...

When a digital forensics investigator suspects that steganography has been used to hide data in an image, he must not only determine that the image contains embedded information but also identify the method used for embedding. The determination of the embedding method – or stego fingerprint – is critical to extracting the hidden information. This paper focuses on identifying stego fingerprints ...

We study the topmost weight matrix of neural network language models. We show that this matrix constitutes a valid word embedding. When training language models, we recommend tying the input embedding and this output embedding. We analyze the resulting update rules and show that the tied embedding evolves in a more similar way to the output embedding than to the input embedding in the untied mo...

ABSRACT The usual approach to nonlinear time series modeling consists of two steps: reconstruct the topological structure of the dynamics (with a time delay embedding), and perform some surface fitting to estimate the evolution operator. We show that the usual method of estimating time delay and embedding dimension is sub-optimal when one is concerned with modeling the dynamics. We introduce an...

This paper proposes an improved watermarking algorithm based on DCT (Discrete Cosine Transform). We carried out the algorithm as described as follows. First, we extended both rows and ranks of the watermark by using the proposed method before the embedding stage. After expansion, Sine chaotic system is employed in encrypting the watermark. In the embedding stage, an effective and adaptive embed...