This paper investigates audio-visual interaction, i.e. inter-modal influences, in linear-regressive model adaptation for multi-modal speech recognition. In the multi-modal adaptation, inter-modal information may contribute the performance of speech recognition. Thus the influence and advantage of intermodal elements should be examined. Experiments were conducted to evaluate several transformati...

We consider the problem of duplicate detection: given a large data set in which each entry has multiple attributes, detect which distinct entries refer to the same real world entity. Our method consists of three main steps: creating a similarity score between entries, grouping entries together into ‘unique entities’, and refining the groups. We compare various methods for creating similarity sc...

We present an extension to the Parametrized Self-Organizing Map that allows the construction of continuous manifolds from noisy, incomplete and not necessarily gridorganized training data. All three problems are tackled by minimizing the overall smoothness of a PSOM manifold. For this, we introduce a matrix which defines a metric in the space of PSOM weights, depending only on the underlying gr...

We discuss the ramifications of noisy and incomplete observations of network data on the existence of a giant connected component (GCC). The existence of a GCC in a random graph can be described in terms of a percolation process, and building on general results for classes of random graphs with specified degree distributions we derive percolation thresholds above which GCCs exist. We show that ...

We propose three novel methods for recovering edges in piecewise smooth functions from their possibly incomplete and noisy spectral information. The proposed methods utilize three different approaches: #1. the randomly-based sparse Inverse Fast Fourier Transform (sIFT); #2. the Total Variation-based (TV) compressive sensing; and #3. the modified zero crossing. The different approaches share a c...

We propose here a nev method for retrieving an image from noisy, incomplete data. It is based on an idea incorporating the principle of maximum entropy, a veil known powerful method for image processing. An image is built on the computer by allocating grains to pixels one at a time on the basis of certain deterministic rules as to where each successive grain is placed in image space. The perfor...

Owing to the robustness of large sparse corruptions and the discrimination of class labels, sparse signal representation has been one of the most advanced techniques in the fields of pattern classification, computer vision, machine learning and so on. This paper investigates the problem of robust face classification when a test sample has missing values. Firstly, we propose a classification met...