Hagai Aronowitz

Fast High Dimensional Vector Multiplication Face Recognition

This paper advances descriptor-based face recognition by suggesting a novel usage of descriptors to form an over-complete representation, and by proposing a new metric learning pipeline within the same/not-same framework. First, the Over-Complete Local Binary Patterns (OCLBP) face representation scheme is introduced as a multi-scale modified version of the Local Binary Patterns (LBP) scheme. Second, we propose an efficient matrix-vector multiplication-based recognition system. The system is based on Linear Discriminant Analysis (LDA) coupled with Within Class Covariance Normalization (WCCN). This is further extended to the unsupervised case by proposing an unsupervised variant of WCCN. Lastly, we introduce Diffusion Maps (DM) for non-linear dimensionality reduction as an alternative to the Whitened Principal Component Analysis (WPCA) method which is often used in face recognition. We evaluate the proposed framework on the LFW face recognition dataset under the restricted, unrestricted and unsupervised protocols. In all three cases we achieve very competitive results.

Efficient Speaker Recognition Using Approximated Cross Entropy (ACE)

Techniques for efficient speaker recognition are presented. These techniques are based on approximating Gaussian mixture modeling (GMM) likelihood scoring using approximated cross entropy (ACE). Gaussian mixture modeling is used for representing both training and test sessions and is shown to perform speaker recognition and retrieval extremely efficiently without any notable degradation in accuracy compared to classic GMM-based recognition. In addition, a GMM compression algorithm is presented. This algorithm decreases considerably the storage needed for speaker retrieval.

Speaker indexing in audio archives using gaussian mixture scoring simulation

Speaker indexing has recently emerged as an important task due to the rapidly growing volume of audio archives. Current filtration techniques still suffer from problems both in accuracy and efficiency. In this paper an efficient method to simulate GMM scoring is presented. Simulation is done by fitting a GMM not only to every target speaker but also to every test utterance, and then computing the likelihood of the test call using these GMMs instead of using the original data. GMM simulation is used to achieve very efficient speaker indexing in terms of both search time and index size. Results on the SPIDRE and NIST-2004 speaker evaluation corpuses show that our approach maintains and sometimes exceeds the accuracy of the conventional GMM algorithm and achieves efficient indexing capabilities: 6000 times faster than a conventional GMM with 1% overhead in storage.

A session-GMM generative model using test utterance Gaussian mixture modeling for speaker verification

Test utterance parameterization (TUP) using Gaussian mixture models (GMMs) has recently been shown to be beneficial for speaker indexing due to its computational efficiency and identical accuracy compared to classic GMM-based recognizers. We show that TUP can also lead to more accurate speaker recognition. On the NIST-2004 evaluation corpus, recognition error rate was reduced by 8% compared to the classic GMM-based algorithm. Furthermore, we introduce a novel generative statistical model for generation of test utterances by speakers. This model is incorporated naturally into the TUP framework and improves speaker recognition accuracy. On the NIST-2004 evaluation corpus, recognition error rate was reduced by 15% compared to the classic GMM-based algorithm.

Text-Dependent Audiovisual Synchrony Detection for Spoofing Detection in Mobile Person Recognition.

Liveness detection is an important countermeasure against spoofing attacks on biometric authentication systems. In the context of audiovisual biometrics, synchrony detection is a proposed method for liveness confirmation. This paper presents a novel, text-dependent scheme for checking audiovisual synchronization in a video sequence. We present custom visual features learned using a unique deep learning framework and show that they outperform other commonly used visual features. We tested our system on two testing sets representing realistic spoofing attack approaches. On our mobile dataset of short video clips of people talking, we obtained equal error rates of 0.8% and 2.7% for liveness detection of photos and video attacks, respectively.

Reducing Noise Bias in the i-Vector Space for Speaker Recognition.

In this paper we develop a simple mathematical model for reducing speaker recognition noise bias in the i-vector space. The method was successfully tested on two different databases covering distinct microphones and background noise scenarios. Substantial reduction in score variability was attained across distinct evaluation conditions which is particularly important in forensic applications. Although originally designed for addressing additive noise, we show that under certain circumstances the proposed method incidentally alleviates convolutive nuisance as well.

Speech processing and retrieval in a personal memory aid system for the elderly

The paper presents a new application of automatic speech processing in the Ambient Assisted Living area, developed in the course of a three year research project. Recording and automatic processing of spoken conversations plays a major role in this solution enabling effective search in a personal audio archive and fast browsing of conversations. Processing of elderly conversational speech recorded by a distant PDA microphone poses a great challenge. The speech processing flow includes transcription, speaker tracking and combined indexing and search of spoken terms and participating speakers identity extracted from the audio. We present the entire application and individual speech processing components as well as evaluation results of the individual components and of the end-to-end spoken information retrieval solution.