In this paper, we consider the task of language identification in the context of mismatch conditions. Specifically, we address the issue of using unlabeled data in the domain of interest to improve the performance of a state-of-the-art system. The evaluation is performed on a 9-language set that includes data in both conversational telephone speech and narrowband broadcast speech. Multiple experiments are conducted to assess the performance of the system in this condition and a number of alternatives to ameliorate the drop in performance. The best system evaluated is based on deep neural network (DNN) bottleneck features using i-vectors utilizing a combination of all the approaches proposed in this work. The resulting system improved baseline DNN system performance by 30%.

Edge detection is an important task in image processing, and the quality of further processing is often reflected by the quality of edge detector outputs. Therefore, it is necessary to develop effective edge map quality measures to assist in evaluating the performance of edge detectors. Objective evaluation measures are crucial in automatically determining the optimal edge map for a given image or an application, as well as its parameter values. In this paper, a new reference-based edge measure (RBEM) is introduced to evaluate the performance of edge detector outputs relative to a ground truth. The new measure fuses four component metrics, based on edge pixel presence, edge corner localization, thick edge occurrence, and edge connectivity. Each of these metrics can be used separately or as a standalone measure to evaluate the quality of an edge map in terms of specific characteristics. The effectiveness of the proposed measure is demonstrated for selecting the best edge detector among several edge detectors, as well as for selecting the optimal parameter values, for both synthetic images and natural images. Experimental results show that the presented RBEM outperforms the existing methods according to subjective evaluation mean opinion scores, as it considers more important visual features in its evaluation.

Image enhancement is a crucial pre-processing step for various image processing applications and vision systems. Many enhancement algorithms have been proposed based on different sets of criteria. However, a direct multi-scale image enhancement algorithm capable of independently and/or simultaneously providing adequate contrast enhancement, tonal rendition, dynamic range compression, and accurate edge preservation in a controlled manner has yet to be produced. In this paper, a multi-scale image enhancement algorithm based on a new parametric contrast measure is presented. The parametric contrast measure incorporates not only the luminance masking characteristic, but also the contrast masking characteristic of the human visual system. The formulation of the contrast measure can be adapted for any multi-resolution decomposition scheme in order to yield new human visual system-inspired multi-scale transforms. In this article, it is exemplified using the Laplacian pyramid, discrete wavelet transform, stationary wavelet transform, and dual-tree complex wavelet transform. Consequently, the proposed enhancement procedure is developed. The advantages of the proposed method include: 1) the integration of both the luminance and contrast masking phenomena; 2) the extension of non-linear mapping schemes to human visual system inspired multi-scale contrast coefficients; 3) the extension of human visual system-based image enhancement approaches to the stationary and dual-tree complex wavelet transforms, and a direct means of; 4) adjusting overall brightness; and 5) achieving dynamic range compression for image enhancement within a direct multi-scale enhancement framework. Experimental results demonstrate the ability of the proposed algorithm to achieve simultaneous local and global enhancements.

Image fusion algorithms combine images obtained using different sensors into a single image to provide contextual enhancement of the scene being observed. The fusion of images obtained using infrared (IR) and visible light (VL) cameras are particularly appealing for security applications such as surveillance and concealed weapon detection. In this paper, a new image fusion algorithm is proposed, which considers both the luminance masking feature of the human visual system (HVS) and the nature of the pertinent information in IR images in the context of surveillance applications. Experimental results illustrate the improved performance of the proposed algorithm by both qualitative and quantitative means.

Image fusion is the process of combining multiple images into a single image which retains the most pertinent information from each original image source. More recently, multi-scale image fusion approaches have emerged as a means of providing a more meaningful fusion which better reflects the human visual system. In this paper, multi-scale decomposition techniques and image fusion algorithms are adapted using the Parameterized Logarithmic Image Processing (PLIP) model, a nonlinear image processing framework which more accurately processes images. Experimental results via computer simulations illustrate the improved performance of the proposed algorithms by both qualitative and quantitative means.

This paper introduces a new concept of Boolean derivatives as a fusion of partial derivatives of Boolean functions (PDBFs). Three efficient algorithms for the calculation of PDBFs are presented. It is shown that Boolean function derivatives are useful for the application of identifying the location of edge pixels in binary images. The same concept is extended to the development of a new edge detection algorithm for grayscale images, which yields competitive results, compared with those of traditional methods. Furthermore, a new measure is introduced to automatically determine the parameter values used in the thresholding portion of the binarization procedure. Through computer simulations, demonstrations of Boolean derivatives and the effectiveness of the presented edge detection algorithm, compared with traditional edge detection algorithms, are shown using several synthetic and natural test images. In order to make quantitative comparisons, two quantitative measures are used: one based on the recovery of the original image from the output edge map and the Pratt's figure of merit.

Edge detection has been used extensively as a preprocessing step for many computer vision tasks. Due to its importance in image processing and the highly subjective nature of human evaluation and visual comparison of edge detectors, it is desirable to formulate objective edge map evaluation measures. One would like to use such a measure to make comparisons of results using the same edge detector with different parameters as well as to make comparisons of results using different edge detectors. Reconstruction-based measures have the clear advantage that they effectively incorporate original image data. In this paper, a general model for reconstruction-based measures is established in order to alleviate the shortcomings of the reconstruction-based measures, followed by the formulation of a new non-reference measure for objective edge map evaluation. Experimental results illustrate the effectiveness of the new measure both as a means of selecting optimal edge detector parameters and as a means of determining the relative performance of edge detectors for a given image.

Feature extraction is arguably the most important stage of an automatic object detection system. It is in this stage where the results of previous processing steps are interpreted to somehow characterize an object. Developing methods for feature extraction and feature vector generation using information from edge maps is a natural progression, as edge detection determines structure in images. A new edge-based feature extraction scheme is introduced based on the feature fusion of two existing methods. A generalized set of kernels for edge detection is also presented. The experimental results show that the detection of different objects of interests is improved using the new method.