but effective GMMs for likelihood functions, a universal background model (UBM) for alternative speaker representation, and a form of Bayesian adaptation to derive speaker models from the UBM. The development and use of a handset detector and score normalization to greatly improve verification performance

The problem of efficient, interactive foreground/background segmentation in still images is of great practical importance in image editing. Classical image segmentation tools use either texture (colour) information, e.g. Magic Wand, or edge (contrast) information, e.g. Intelligent Scissors. Recently, an approach based on optimization by graph-cut has been developed which successfully combines both types of information. In this paper we extend the graph-cut approach in three respects. First, we have developed a more powerful, iterative version of the optimisation. Secondly, the power of the iterative algorithm is used to simplify substantially the user interaction needed for a given quality of result. Thirdly, a robust algorithm for “border matting ” has been developed to estimate simultaneously the alpha-matte around an object boundary and the colours of foreground pixels. We show that for moderately difficult examples the proposed method outperforms competitive tools.

Abstract — An established approach to training Gaussian Mixture Models (GMMs) for speaker verification is via the expectation-maximisation (EM) algorithm. The EM algorithm has been shown to be sensitive to initialisation and prone to converging on local maxima. In exploration of these issues, three

We report on the construction of a real-time computer system capable of distinguishing speech signals from music signals over a wide range of digital audio input. We have examined 13 features intended to measure conceptually distinct properties of speech and/or music signals, and combined them in several multidimensional classification frameworks. We provide extensive data on systemperformanceand the cross-validated training/test setup used to evaluate the system. For the datasets currently in use, the best classifier classifies with 5.8% error on a frame-by-frame basis, and 1.4% error when integrating long (2.4 second) segments of sound. 1. OVERVIEW The problem of distinguishing speech signals from music signals has become increasingly important as automatic speech recognition (ASR) systems are applied to more and more "real-world" multimedia domains. If we wish to build systems that perform ASR on soundtrack data, for example, it is important to be able to distinguish which segments...

The Fisher kernel (FK) is a generic framework which combines the benefits of generative and discriminative approaches. In the context of image classification the FK was shown to extend the popular bag-of-visual-words (BOV) by going beyond count statistics. However, in practice, this enriched representation has not yet shown its superiority over the BOV. In the first part we show that with several well-motivated modifications over the original framework we can boost the accuracy of the FK. On PASCAL VOC 2007 we increase the Average Precision (AP) from 47.9 % to 58.3%. Similarly, we demonstrate state-of-the-art accuracy on CalTech 256. A major advantage is that these results are obtained using only SIFT descriptors and costless linear classifiers. Equipped with this representation, we can now explore image classification on a larger scale. In the second part, as an application, we compare two abundant resources of labeled images to learn classifiers: ImageNet and Flickr groups. In an evaluation involving hundreds of thousands of training images we show that classifiers learned on Flickr groups perform surprisingly well (although they were not intended for this purpose) and that they can complement classifiers learned on more carefully annotated datasets.

mixture models, GMMs, on the frame level. The method has been tested on two different corpora and languages; Swedish voice controlled telephone services and English meetings. The results indicate that using GMMs on the frame level is a feasible technique for emotion classification. The two MFCC methods

is two fold: i) an optimally compact visual dictionary is learned by pair-wise merging of visual words from an initially large dictionary. The final visual words are described by GMMs. ii) A novel statistical measure of discrimination is proposed which is optimized by each merge operation. High

This paper compares three approaches to building phoneme-specific Gaussian mixture model (GMM) speaker recognition systems on the NIST 2003 Extended Data Evaluation to a baseline GMM system covering all of the phonemes. The individual performance of any given phoneme-specific GMM system falls below the performance of the baseline GMM, but fusing the top 40 performing scores of the individual phoneme systems at the 8 conversation train condition resulted in an equal error rate of 1.7%, which is a 2.6 % absolute reduction in equal error rate from the baseline system. Further investigation showed complementary information across the three model building approaches as error rates dropped on a per phoneme basis when these systems were fused. 1.

In this paper we present an adapted UBM-GMM based privacy preserving speaker verification (PPSV) system, where the system is not able to observe the speech data provided by the user and the user does not observe the models trained by the system. These privacy criteria are important in order to prevent an adversary having unauthorized access to the user’s client device from impersonating a user and also from another adversary who can break into the verification system can learn about the user’s speech patterns to impersonate the user in another system. We present protocols for speaker enrollment and verification which preserve privacy according to these requirements and report experiments with a prototype implementation on the YOHO dataset. Index Terms: speaker verification, secure biometrics 1.

of prior information and uncertainty of parameter estimates into the scoring process. A novel development of Bayes factors for GMMs is presented based on incremental adaptation that is well-suited to inclusion in existing state-of-the-art GMM-UBM systems. Experiments on the 1999 NIST Speaker Recognition