Cross likelihood ratio based speaker clustering using eigenvoice models

This paper proposes the use of eigenvoice modeling techniques with the Cross Likelihood Ratio (CLR) as a criterion for speaker clustering within a speaker diarization system. The CLR has previously been shown to be a robust decision criterion for speaker clustering using Gaussian Mixture Models. Recently, eigenvoice modeling techniques have become increasingly popular, due to its ability to adequately represent a speaker based on sparse training data, as well as an improved capture of differences in speaker characteristics. This paper hence proposes that it would be beneficial to capitalize on the advantages of eigenvoice modeling in a CLR framework. Results obtained on the 2002 Rich Transcription (RT-02) Evaluation dataset show an improved clustering performance, resulting in a 35.1% relative improvement in the overall Diarization Error Rate (DER) compared to the baseline system.

Multiple cameras for audio-visual speech recognition in an automotive environment

Audio-visualspeechrecognition, or the combination of visual lip-reading with traditional acoustic speechrecognition, has been previously shown to provide a considerable improvement over acoustic-only approaches in noisy environments, such as that present in an automotive cabin. The research presented in this paper will extend upon the established audio-visualspeechrecognition literature to show that further improvements in speechrecognition accuracy can be obtained when multiple frontal or near-frontal views of a speaker's face are available. A series of visualspeechrecognition experiments using a four-stream visual synchronous hidden Markov model (SHMM) are conducted on the four-camera AVICAR automotiveaudio-visualspeech database. We study the relative contribution between the side and central orientated cameras in improving visualspeechrecognition accuracy. Finally combination of the four visual streams with a single audio stream in a five-stream SHMM demonstrates a relative improvement of over 56% in word recognition accuracy when compared to the acoustic-only approach in the noisiest conditions of the AVICAR database.

Speaker diarization : "who spoke when"

Speaker diarization is the process of annotating an input audio with information that attributes temporal regions of the audio signal to their respective sources, which may include both speech and non-speech events. For speech regions, the diarization system also specifies the locations of speaker boundaries and assign relative speaker labels to each homogeneous segment of speech. In short, speaker diarization systems effectively answer the question of ‘who spoke when’. There are several important applications for speaker diarization technology, such as facilitating speaker indexing systems to allow users to directly access the relevant segments of interest within a given audio, and assisting with other downstream processes such as summarizing and parsing. When combined with automatic speech recognition (ASR) systems, the metadata extracted from a speaker diarization system can provide complementary information for ASR transcripts including the location of speaker turns and relative speaker segment labels, making the transcripts more readable. Speaker diarization output can also be used to localize the instances of specific speakers to pool data for model adaptation, which in turn boosts transcription accuracies. Speaker diarization therefore plays an important role as a preliminary step in automatic transcription of audio data. The aim of this work is to improve the usefulness and practicality of speaker diarization technology, through the reduction of diarization error rates. In particular, this research is focused on the segmentation and clustering stages within a diarization system. Although particular emphasis is placed on the broadcast news audio domain and systems developed throughout this work are also trained and tested on broadcast news data, the techniques proposed in this dissertation are also applicable to other domains including telephone conversations and meetings audio. Three main research themes were pursued: heuristic rules for speaker segmentation, modelling uncertainty in speaker model estimates, and modelling uncertainty in eigenvoice speaker modelling. The use of heuristic approaches for the speaker segmentation task was first investigated, with emphasis placed on minimizing missed boundary detections. A set of heuristic rules was proposed, to govern the detection and heuristic selection of candidate speaker segment boundaries. A second pass, using the same heuristic algorithm with a smaller window, was also proposed with the aim of improving detection of boundaries around short speaker segments. Compared to single threshold based methods, the proposed heuristic approach was shown to provide improved segmentation performance, leading to a reduction in the overall diarization error rate. Methods to model the uncertainty in speaker model estimates were developed, to address the difficulties associated with making segmentation and clustering decisions with limited data in the speaker segments. The Bayes factor, derived specifically for multivariate Gaussian speaker modelling, was introduced to account for the uncertainty of the speaker model estimates. The use of the Bayes factor also enabled the incorporation of prior information regarding the audio to aid segmentation and clustering decisions. The idea of modelling uncertainty in speaker model estimates was also extended to the eigenvoice speaker modelling framework for the speaker clustering task. Building on the application of Bayesian approaches to the speaker diarization problem, the proposed approach takes into account the uncertainty associated with the explicit estimation of the speaker factors. The proposed decision criteria, based on Bayesian theory, was shown to generally outperform their non- Bayesian counterparts.

Robust automatic speaker linking and attribution

This research makes a major contribution which enables efficient searching and indexing of large archives of spoken audio based on speaker identity. It introduces a novel technique dubbed as “speaker attribution” which is the task of automatically determining ‘who spoke when?’ in recordings and then automatically linking the unique speaker identities within each recording across multiple recordings. The outcome of the research will also have significant impact in improving the performance of automatic speech recognition systems through the extracted speaker identities.

Cross database training of audio-visual hidden Markov models for phone recognition

Speech recognition can be improved by using visual information in the form of lip movements of the speaker in addition to audio information. To date, state-of-the-art techniques for audio-visual speech recognition continue to use audio and visual data of the same database for training their models. In this paper, we present a new approach to make use of one modality of an external dataset in addition to a given audio-visual dataset. By so doing, it is possible to create more powerful models from other extensive audio-only databases and adapt them on our comparatively smaller multi-stream databases. Results show that the presented approach outperforms the widely adopted synchronous hidden Markov models (HMM) trained jointly on audio and visual data of a given audio-visual database for phone recognition by 29% relative. It also outperforms the external audio models trained on extensive external audio datasets and also internal audio models by 5.5% and 46% relative respectively. We also show that the proposed approach is beneficial in noisy environments where the audio source is affected by the environmental noise.

A comparative study of speaker adaptation methods

For the problem of speaker adaptation in speech recognition, the performance depends on the availability of adaptation data. In this paper, we have compared several existing speaker adaptation methods, viz. maximum likelihood linear regression (MLLR), eigenvoice (EV), eigenspace-based MLLR (EMLLR), segmental eigenvoice (SEV) and hierarchical eigenvoice (HEV) based methods. We also develop a new method by modifying the existing HEV method for achieving further performance improvement in a limited available data scenario. In the sense of availability of adaptation data, the new modified HEV (MHEV) method is shown to perform better than all the existing methods throughout the range of operation except the case of MLLR at the availability of more adaptation data.

Joint decoding of multiple speech patterns for robust speech recognition

We are addressing a new problem of improving automatic speech recognition performance, given multiple utterances of patterns from the same class. We have formulated the problem of jointly decoding K multiple patterns given a single Hidden Markov Model. It is shown that such a solution is possible by aligning the K patterns using the proposed Multi Pattern Dynamic Time Warping algorithm followed by the Constrained Multi Pattern Viterbi Algorithm The new formulation is tested in the context of speaker independent isolated word recognition for both clean and noisy patterns. When 10 percent of speech is affected by a burst noise at -5 dB Signal to Noise Ratio (local), it is shown that joint decoding using only two noisy patterns reduces the noisy speech recognition error rate to about 51 percent, when compared to the single pattern decoding using the Viterbi Algorithm. In contrast a simple maximization of individual pattern likelihoods, provides only about 7 percent reduction in error rate.

Design of speaker identification schemes for large number of speakers (A)

Design of speaker identification schemes for a small number of speakers (around 10) with a high degree of accuracy in controlled environment is a practical proposition today. When the number of speakers is large (say 50–100), many of these schemes cannot be directly extended, as both recognition error and computation time increase monotonically with population size. The feature selection problem is also complex for such schemes. Though there were earlier attempts to rank order features based on statistical distance measures, it has been observed only recently that the best two independent measurements are not the same as the combination in two's for pattern classification. We propose here a systematic approach to the problem using the decision tree or hierarchical classifier with the following objectives: (1) Design of optimal policy at each node of the tree given the tree structure i.e., the tree skeleton and the features to be used at each node. (2) Determination of the optimal feature measurement and decision policy given only the tree skeleton. Applicability of optimization procedures such as dynamic programming in the design of such trees is studied. The experimental results deal with the design of a 50 speaker identification scheme based on this approach.

Feature Selection for Speech Emotion Recognition in Spanish and Basque: On the Use of Machine Learning to Improve Human-Computer Interaction

Study of emotions in human-computer interaction is a growing research area. This paper shows an attempt to select the most significant features for emotion recognition in spoken Basque and Spanish Languages using different methods for feature selection. RekEmozio database was used as the experimental data set. Several Machine Learning paradigms were used for the emotion classification task. Experiments were executed in three phases, using different sets of features as classification variables in each phase. Moreover, feature subset selection was applied at each phase in order to seek for the most relevant feature subset. The three phases approach was selected to check the validity of the proposed approach. Achieved results show that an instance-based learning algorithm using feature subset selection techniques based on evolutionary algorithms is the best Machine Learning paradigm in automatic emotion recognition, with all different feature sets, obtaining a mean of 80,05% emotion recognition rate in Basque and a 74,82% in Spanish. In order to check the goodness of the proposed process, a greedy searching approach (FSS-Forward) has been applied and a comparison between them is provided. Based on achieved results, a set of most relevant non-speaker dependent features is proposed for both languages and new perspectives are suggested.

Modified Kanerva model. Results for real time word recognition

This paper describes results obtained using the modified Kanerva model to perform word recognition in continuous speech after being trained on the multi-speaker Alvey 'Hotel' speech corpus. Theoretical discoveries have recently enabled us to increase the speed of execution of part of the model by two orders of magnitude over that previously reported by Prager & Fallside. The memory required for the operation of the model has been similarly reduced. The recognition accuracy reaches 95% without syntactic constraints when tested on different data from seven trained speakers. Real time simulation of a model with 9,734 active units is now possible in both training and recognition modes using the Alvey PARSIFAL transputer array. The modified Kanerva model is a static network consisting of a fixed nonlinear mapping (location matching) followed by a single layer of conventional adaptive links. A section of preprocessed speech is transformed by the non-linear mapping to a high dimensional representation. From this intermediate representation a simple linear mapping is able to perform complex pattern discrimination to form the output, indicating the nature of the speech features present in the input window.

Continuous speech recognition for the TIMIT database using neural networks

Four types of neural networks which have previously been established for speech recognition and tested on a small, seven-speaker, 100-sentence database are applied to the TIMIT database. The networks are a recurrent network phoneme recognizer, a modified Kanerva model morph recognizer, a compositional representation phoneme-to-word recognizer, and a modified Kanerva model morph-to-word recognizer. The major result is for the recurrent net, giving a phoneme recognition accuracy of 57% from the si and sx sentences. The Kanerva morph recognizer achieves 66.2% accuracy for a small subset of the sa and sx sentences. The results for the word recognizers are incomplete.