Colloque avec actes et comité de lecture.

We describe a continuous speech recognition system that uses the multi-band paradigm. This principle is based on the recombination of several independent sub-recognizers, each one assigned to a specific frequency band. The major issue of such systems consists of deciding at which time the recombination must be done. Our algorithm lets each band be totally independent from the others, and uses the different solutions to resegment the initial sentence. Finally, the bands are synchronously merged together, according to this new segmentation. The whole system is too complex to be entirely described here, and, in this paper, we concentrate on the synchronous recombination part, which is achieved by a classifier. The system has been tested in clean and noisy environments, and proved to be especially robust to noise.

Second order hidden Markov models have been used for a long time in pattern recognition, especially in speech recognition. Their main advantages over other methods (neural networks etc.) are their capabilities of modeling noisy temporal signals of variable length. In a previous work, we proposed a new method based on second order hidden Markov models to learn and recognize places in an indoor environment by a mobile robot, and showed that this approach is well suited for learning and recognizing places. In this paper, we propose major modifications to increase the global rate of place recognition. Results of experiments on a real robot with distinctive places are given.

The problem addressed by this paper is to enhance the continuous speech recognizers robustness to noise. For this purpose, the acoustic signal is filtered into several spectral bands, and independent recognition is achieved in each band. Then, the system recombines the results given by each recognizer and delivers a unique solution. The main advantage of this method is to consider the signal only in the bands which are relevant, and to ignore spectral bands which are corrupted by noise. We are developping a speaker-independent continuous speech recognizer based on this principle.

The phonetic context has a large effect on phonemes in a continuous speech signal [1]. Therefore recognition systems that model allophones using context-dependent Hidden Markov Models have been implemented [4]. Second-order HMMs (HMM2s have a great ability for the segmentation in the temporal domain [6][7] but have some difficulties in the recognition because the MLE training (Maximum Likelihood Estimation) is not discriminant, whereas the discrimination is one of the abilities of the Artificial Neural Networks models. In the last three years we have developed a new ANN model named OWE (Orthogonal Weight Estimator)[10][11]. The principle of the OWE is a ANN that classifies an input pattern according to contextual environment. This new ANN architecture tackles the problem of context dependent behaviour training. Roughly, the principle is based on main MLP (Multilayered Perceptron) in which each synaptic weight connection value is estimated by another MLP (an OWE) with respect to context representation. In this paper, we present 2 hybrid systems for phoneme recognition. In both systems, 48 context independent HMM2s segment the input signal. In the first system, the OWE performs the labelling of segments and, in the second system, the OWE outputs are the input frames of the HMM2s. Experiments on TIMIT range from 56% to 67% accuracies on the 48 phonemes set.

In this paper, we propose a new method based on hidden Markov models to learn and recognize places in an indoor environment by a mobile robot. Hidden Markov models have been used for a long time in pattern recognition, especially in speech recognition. Their main advantages over other methods (e.g. neural networks) are their capabilities to modelize noisy temporal signals of variable length. We show in this paper that this approach is well adapted for learning and recognition of places by a mobile robot. Results of experiments on a real robot with five distinctive places are given.

We propose an extension of the Viterbi algorithm that makes second-order hidden Markov models computationally efficient. A comparative study between first-order (HMM1s) and second-order Markov models (HMM2s) is carried out. Experimental results show that HMM2s provide a better state occupancy modeling and, alone, have performances comparable with HMM1s plus postprocessing.

In the field of speech recognition by stochastic methods, it is conventional to pursue approaches using first-order-hidden Markov models (HMM1s). Despite the success of this approach, it is still worth investigating if some of the drawbacks of HMM1s can be overcome, e.g. by using higher-order Markov processes. In this paper, we show that second-order hidden Markov models (HMM2s) can yield high performances in the context of continuous speech recognition. We first present the underlying equations and complexity of HMM2s in the maximum likelihood estimation (MLE) paradigm. Then, we show that in a connected word recognition task, such as spelled name recognition over the telephone, HMM2s outperform HMM1s. In the field of phoneme-based continuous speech recognition, we show that context-independent HMM2s can achieve more than 69% phone accuracy.

The phonetic context has a large effect on stop consonants in a continuous speech signal [1].Therefore recognition systems that model allophones using context-dependent Hidden Markov Models have been implemented [3].HMMs have a great ability for the segmentation in the temporal domain [4][6] but have some difficulties in the recognition because the MLE training (Maximum Likelihood Estimation) is not discriminant, whereas the discrimination is one of the abilities of the Artificial Neural Networks models.In the last three years we have developed a new ANN model named OWE (Orthogonal Weight Estimator) [9][10].The principle of the OWE is a ANN that classifies an input pattern according to contextual environment.This new ANN architecture tackles the problem of context dependent behaviour training.Roughly, the principle is based on main MLP (Multilayered Perceptron) in which each synaptic weight connection value is estimated by another MLP (an OWE) with respect to context representation.In this paper, we present a hierarchical system for phoneme recognition: first the system segments the input signal using 48 context independent HMMs.Then the stop consonant are reordered by a OWE ANN.Experiments on TIMIT show 78 % of correct recognition rate on the 6 stop consonants (/p, t, k, b, d, g).

We introduce SmarTspelL, a new speaker-independent algorithm to recognize continuously spelled names over the telephone. Our method is based on an N-best multi-pass recognition strategy applying costly constraints when the number of possible candidates is low. This strategy outperforms an HMM recognizer using a grammar containing all the possible names. It is also more suitable to real-time implementation. For a 3388 name dictionary, a 95.3% name recognition rate is obtained. A real-time prototype has been implemented on a workstation. We also present comparisons of different feature sets for speech representation, and two speech recognition approaches based on first- and second-order HMMs.

Representation of speech using hidden Markov models is compared with a representation based on finite-state automata. Experimental results are presented for both methods and the respective advantages and disadvantages discussed.>

This paper presents a new method for connected-word recognition with confusable vocabularies, such as connected letters. The recognition process is performed in two steps. First, a second-order HMM provides N-best word strings. Then, the strings of confusable letters are discriminated by a procedure based on acoustic knowledge and artificial neural networks (ANN). This method has been tested on an American-English database containing spelled names collected through the telephone network. The results obtained with the first HMM pass and the improvements made with the ANN are presented and discussed. When a 3,300 name dictionary and a retrieval procedure based on a DTW alignment algorithm were used, 96% recognition accuracy was obtained. I. INTRODUCTION The performances of HMM recognizers are now quite satisfactory for small vocabularies. However, in the case of confusable words, results are not sufficient for real-world applications, especially in adverse conditions like noisy or tele...

A recognizer for isolated words based on second order hidden Markov models (HMMs) is presented. This system is compared with the usual first order one. To evaluate the performances of both systems, a database of French digits is used. Results are more accurate with the recognizer using second-order hidden Markov process. A version of the Viterbi algorithm which is suited to continuous-speech recognition is described. This version, called Viterbi-block allows segmentation and labeling of an unknown utterance during its processing, whereas the classical Viterbi algorithm waits until the end of the sentence before backtracking the best path. This algorithm is based on a local optimum comparison of the different probabilities computed by the Viterbi algorithm during the time warping of a shift window of fixed length in the signal and the different HMMs. Different criteria of comparison between the probabilities of different paths are tried and assessed.>

We present in this paper a new approach to acoustic-phonetic decoding of continuous speech that consists of integrating the two most promising present techniques in the field, i.e. stochastic modeling and knowledge-based expert systems. Our group has been developing several systems based on these two techniques during the past 15 years or so. Our present goal is to mix both in order to improve the overall quality of automatic phonetic decoding. This paper is concerned with the first developments of the project. We first present the main characteristics of the components of the system and its general architecture. We also present and discuss preliminary results concerning the segmentation of the speech wave into phonetic units and the gross labeling of these segments.

Our group has been designing for the past twelve years several speech recognition systems, from isolated vocabulary pattern matching systems to continuous speech understanding systems. The experiments we carried out showed us that the systems designed for restricted vocabularies task were not readily extensible to large vocabularies. We therefore started some years ago implementing a 200 word recognition system using a phonetic approach. This system was tested successfully in 1980. In continuation of this research we decided to extend our approach to a 1000 word vocabulary. This paper describes the principles involved in this system together with the preliminary results already obtained. The basic idea is to reduce the number of word candidates for the recognition by looking for robust phonetic features computed from the input signal. These features are used as a key for accessing the lexicon. Since the determination of the features is carried out in parallel with the phonetic decoding of the input word, it is possible to design a multiprocessor structure in order to reduce the overall recognition time. The determination of crude phonetic features is described together with the organization of the lexicon. Some preliminary results are finally presented and discussed.