Search

In this article, we are interested in spoken term detection task, with a particular focus on Person Name (PN) spotting in automatic speech recognition (ASR) system outputs. We propose a two-step method that combines an acoustic matching based on a Phoneme Confusion Network (PCN) with a semantic rescoring based on the Latent Dirichlet Allocation (LDA) models. The first module allows to find, in the PCN, potential PN candidates in speech segments, while the second is in charge of ranking the competing PN, according to a LDA topic model. The proposed LDA-based approach outperforms significantly the baseline system based on a search in the ASR phoneme lattice, obtaining a F-measure score of 77.04% on PN detection.

International audience; —We present a method to detect social events in a set of pictures from an image hosting service (Flickr). This method relies on the analysis of user-generated tags, by using statistical models trained on both a small set of manually annotated data and a large data set collected from the Internet. Social event modeling relies on multi-span topic model based on LDA (Latent Dirichlet Allocation). Experiments are conducted in the experimental setup of MediaEval'2011 evaluation campaign. The proposed system outperforms significantly the best system of this benchmark, reaching a F-measure score of about 71%.

International audience; The detection and characterization, in audiovisual documents, of speech utterances where person names are pronounced, is an important cue for spoken content analysis. This paper tackles the problematic of retrieving spoken person names in the 1-Best ASR outputs of broadcast TV shows. Our assumption is that a person name is a latent variable produced by the lexical context it appears in. Thereby, a spoken name could be derived from ASR outputs even if it has not been proposed by the speech recognition system. A new context modelling is proposed in order to capture lexical and structural information surrounding a spoken name. The fundamental hypothesis of this study has been validated on broadcast TV documents available in the context of the REPERE challenge.

International audience; Combining automatic speech recognition (ASR) systems generally relies on the posterior merging of the outputs or on acoustic cross-adaptation. In this paper, we propose an integrated approach where outputs of secondary systems are integrated in the search algorithm of a primary one. In this driven decoding algorithm (DDA), the secondary systems are viewed as observation sources that should be evaluated and combined to others by a primary search algorithm. DDA is evaluated on a subset of the ESTER I corpus consisting of 4 hours of French radio broadcast news. Results demonstrate DDA significantly outperforms vote-based approaches: we obtain an improvement of 14.5% relative word error rate over the best single-systems, as opposed to the the 6.7% with a ROVER combination. An in-depth analysis of the DDA shows its ability to improve robustness (gains are greater in adverse conditions) and a relatively low dependency on the search algorithm. The application of DDA to both A* and beam-search-based decoder yields similar performances.

This paper explores a novel method for context-dependent models in automatic speech recognition (ASR), in the context of under-resourced languages. We present a simple way to realize a tying states approach, based on a new vectorial representation of the HMM states. This vectorial representation is considered as a vector of a low number of parameters obtained by the Subspace Gaussian Mixture Models paradigm (SGMM). The proposed method does not require phonetic knowledge or a large amount of data, which represent the major problems of acoustic modeling for under-resourced languages. This paper shows how this representation can be obtained and used for tying states. Our experiments, applied on Vietnamese, show that this approach achieves a stable gain compared to the classical approach which is based on decision trees. Furthermore, this method appears to be portable to other languages, as shown in the preliminary study conducted on Berber.

In this paper we propose a new feature normalization based on Factor Analysis (FA) for the problem of acoustic variability in Automatic Speech Recognition (ASR). The FA paradigm was previously used in the field of ASR, in order to model the usefull information: the HMM state dependent acoustic information. In this paper, we propose to use the FA paradigm to model the useless information (speaker- or channel-variability) in order to remove it from acoustic data frames. The transformed training data frames are then used to train new HMM models using the standard training algorithm. The transformation is also applied to the test data before the decoding process. With this approach we obtain, on french broadcast news, an absolute WER reduction of 1.3%.

International audience; A research on the analysis of human/human conversations in a call centre is described. The purpose of the research is to provide short reports of each conversation with information useful for monitoring the call centre efficiency. Data from real users discussing over the telephone with agents are processed by an automatic speech recognition (ASR) system. Reports are grouped into classes by the agents based on predefined taxonomy. A train set of manually transcribed data is used for training the extraction of features relevant to the application and the classification of the conversations. The use of all the words of the application vocabulary, of automatically selected key_words, and of automatically learned sentence chunks containing semantic classes of words are compared and evaluated with a totally different test set. The results show a significant increase in performance when chunks are used even in comparison with the use of bags of words obtained with a boosting algorithm.

—This paper focuses on automatic speech recognition systems combination based on driven decoding paradigms. The driven decoding algorithm (DDA) involves the use of a 1-best hypothesis provided by an auxiliary system as another knowledge source in the search algorithm of a primary system. In previous studies, it was shown that DDA outperforms ROVER when the primary system is guided by a more accurate system. In this paper we propose a new method to manage auxiliary transcriptions which are presented as a bag-of-n-grams (BONG) without temporal matching. These modifications allow to make easier the combination of several hypotheses given by different auxiliary systems. Using BONG combination with hypotheses provided by two auxiliary systems, each of which obtained more than 23% of WER on the same data, our experiments show that a CMU Sphinx based ASR system can reduce its WER from 19.85% to 18.66% which is better than the results reached with DDA or classical ROVER combination.

International audience; This paper presents a semantic confidence measure that aims to predict the relevance of automatic transcripts for a task of Spoken Document Retrieval (SDR). The proposed predicting method relies on the combination of Automatic Speech Recognition (ASR) confidence measure and a Semantic Com-pacity Index (SCI), that estimates the relevance of the words considering the semantic context in which they occurred. Experiments are conducted on the French Broadcast news corpus ESTER, by simulating a classical SDR usage scenario : users submit text-queries to a search engine that is expected to return the most relevant documents regarding the query. Results demonstrate the interest of using semantic level information to predict the transcription indexability.

In this paper, we present a new method for video genre identification based on the linguistic content analysis. This approach relies on the analysis of the most frequent words in the video transcriptions provided by an automatic speech recognition system. Experiments are conducted on a corpus composed of cartoons, movies, news, commercials, documentary, sport and music. On this 7-genre identification task, the proposed transcription-based method obtains up to 80% of correct identification. Finally, this rate is increased to 95% by combining the proposed linguistic-level features with low-level acoustic features.

Video genre identification methods are frequently based on image or motion analysis, which are relatively timeconsuming processes. Since such approaches are tractable by batch processing, as-soon-as-possible identification requires faster methods. In this paper, we investigate the use of audio-only methods for on-the-fly video classification. We propose to use several acoustic feature streams and we evaluate various combination schemes at the frame or at the score level. Results are compared to those obtained by humans, according to the listening duration. Although the system based on model combination slightly outperforms the humans on very soon detection. The latter remain significantly more accurate on long sessions.

Statistical classifiers operate on features that generally include both useful and useless information. These two types of information are difficult to separate in the feature domain. Recently, a new paradigm based on a Latent Factor Analysis (LFA) proposed a model decomposition into usefull and useless components. This method was successfully applied to speaker and language recognition tasks. In this paper, we study the use of LFA for video genre classification by using only the audio channel. We propose a classification method based on short-term cep-stral features and Gaussian Mixture Models (GMM) or Support Vector Machine (SVM) classifiers, that are combined with Factor Analysis (FA). Experiments are conducted on a corpus composed of 5 types of video (musics, commercials, cartoons, movies and news). The relative classification error reduction obtained by using the best factor analysis configuration with respect to the baseline system, Gaussian Mixture Model Universal Background Model (GMM-UBM), is about 56%, corresponding to a correct identification rate of about 90%.

Speech recognition applications are known to require a significant amount of resources. However, embedded speech recognition only authorizes few KB of memory, few MIPS, and small amount of training data. In order to fit the resource constraints of embedded applications, an approach based on a semicontinuous HMM system using state-independent acoustic modelling is proposed. A transformation is computed and applied to the global model in order to obtain each HMM state-dependent probability density functions, authorizing to store only the transformation parameters. This approach is evaluated on two tasks: digit and voice-command recognition. A fast adaptation technique of acoustic models is also proposed. In order to significantly reduce computational costs, the adaptation is performed only on the global model (using related speaker recognition adaptation techniques) with no need for state-dependent data. The whole approach results in a relative gain of more than 20% compared to a basic HMM-based system fitting the constraints.

International audience; — In this paper we present an integrated unsupervised method to produce a quality corpus for training automatic speech recognition system (ASR) using prompts or closed captions. Closed captions and prompts do not always have timestamps and do not necessarily correspond to the exact speech. We propose a method allowing to extract quality corpus from imperfect transcript. The proposed approach works in two steps. During the search, the ASR system finds matching segments in a large prompt database. Matching segments are then used inside a Driven Decoding Algorithm (DDA) to produce a high quality corpus. Results show a F-measure of 96% in term of spotting while the DDA corrects the output according to the prompts: a high quality corpus is easily extracted. 1 Index Terms— speech recognition, closed captioning, corpus building, automatic segmentation I. INTRODUCTION The training of an automatic speech recognition system (ASR) requires large amounts of exact annotated speech. The transcription task is expensive and takes a lot of time. In some situations imperfect transcripts like journalist prompts, closed captions or abstracts are available. This material is available in large quantities. However, these transcripts present two issues: the distance compared to the audio stream and the lack of timestamps. Various approaches propose to use imperfect transcripts for unsupervised ASR training (section II-D). But existing methods are not integrated and have shortcomings: processes are iterative and take a lot of computing time; the lack of times-tamp is forgotten. Moreover existing methods do not use all the potential of imperfect transcripts. The first part of this paper is dedicated to the related work on these issues: the prompts quality, methods to perform ASR alignment with imperfect transcripts, the automatic imperfect transcript segmentation, and finally how to use them for training an ASR system. In a second part, we describe an integrated approach allowing us to solve the two main approximated transcription issues

International audience; Most of the Web-based methods for lexicon augmenting consist in capturing global semantic features of the targeted domain in order to collect relevant documents from the Web. We suggest that the local context of the out-of-vocabulary (OOV) words contains relevant information on the OOV words. With this information, we propose to use the Web to build locally-augmented lexicons which are used in a final local decoding pass. Our experiments confirm the relevance of the Web for the OOV word retrieval. Different methods are proposed to retrieve the hypothesis words. Finally we present the integration of new words in the transcription process based on part-of-speech models. This technique allows to recover 7.6% of the significant OOV words and the accuracy of the system is improved.

In this paper, a new strategy for a fast adaptation of acoustic models is proposed for embedded speech recognition. It relies on a general GMM, which represents the whole acoustic space, associated with a set of HMM state-dependent probability functions modeled as transformations of this GMM. The work presented here takes advantage of this architecture to propose a fast and efficient way to adapt the acoustic models. The adaptation is performed only on the general GMM model, using techniques gathered from the speaker recognition domain. It does not require state-dependent adaptation data and it is very efficient in terms of computational cost. Weevaluate our approach in the voice-command task, using a car-based corpus. This adaptation method achieved a relative error-rate decrease of about 10% even if few adaptation data are available. The complete system allows a total relative gain of more than 20% compared to a basic HMM-based system. Index Terms: speech recognition, compact acoustic models, adaptation

International audience; The combination of Automatic Speech Recognition (ASR) systems generally relies on a posteriori merge of system outputs or on a cross-adaptation. In this paper, we propose an integrated approach where the search of a primary system is driven by the outputs of a secondary one. This method allows to drive the primary system search by using the one-best hypotheses and the word posteriors gathered from the secondary system. Experiments are carried out within the experimental framework of the ESTER evaluation campaign [1]. Results show that the driven decoding algorithm significantly outper-forms the two single ASR systems (-8% of relative WER,-1.7% absolute). Finally, we investigate the interactions between driven decoding and cross-adaptations. The best cross-adaptation strategy in combination with the driven decoding process brings to a final absolute gain of about 1.9% WER.

In this paper a new computation and approximation scheme for Language Model Look-Ahead (LMLA) is introduced. The main benefit of LMLA is sharper pruning of the search space during the LVCSR decoding process. However LMLA comes with its own cost and is known to scale badly with both LM n-gram order and LM size. The proposed method tackles this problem with a divide and conquer approach which enables faster computation without additional WER cost. The obtained results allowed our system to participate in the real-time task of the ESTER Broadcast News transcription evaluation campaign for French.

We present several methods able to fit speech recognition system requirements to cellular phone resources. The proposed techniques are evaluated on a digit recognition task using both French and English corpora. We investigate particularly three aspects of speech processing: acoustic parameterization, recognition algorithms; acoustic modeling. Several parameterization algorithms (LPCC, MFCC and PLP) are compared to the linear predictive coding (LPC) included in the GSM norm. The MFCC and PLP parameterization algorithms perform significantly better than the others. Moreover, feature vector size can be reduced to 6 PLP coefficients, allowing memory and computation resources to be decreased without a significant loss of performance. In order to achieve good performance with reasonable resource needs, we develop several methods to embed a classical HMM-based speech recognition system in a cellular phone. We first propose an automatic on-line building of a phonetic lexicon which allows a minimal but unlimited lexicon. Then we reduce the HMM complexity by decreasing the number of (Gaussian) components per state. Finally, we evaluate our propositions by comparing dynamic time warping (DTW) with our HMM system - in the cellular phone context - for clean conditions. The experiments show that our HMM system outperforms DTW for speaker independent tasks and allows more practical applications for the cellular-phone user interface.

The aim of speaker adaptation techniques is to enhance speaker-independent acoustic models to bring their recognition accuracy as close as possible to the one obtained with speaker-dependent models. Recently, a technique based on a hierarchical structure and the maximum a posteriori criterion was proposed (SMAP) (Shinoda, K. and Lee, C.-H., Proc IEEE ICASSP, 1998). As in SMAP, we assume that the acoustic model parameters are organized in a tree containing all the Gaussian distributions. Each node in that tree represents a cluster of Gaussian distributions sharing a common affine transformation representing the mismatch between training and test conditions. To estimate this affine transformation, we propose a new technique based on merging Gaussians and the standard MAP adaptation. This new technique is very fast and allows a good unsupervised adaptation for both means and variances even with a small amount of adaptation data. This adaptation strategy has shown a significant performance improvement in a large vocabulary speech recognition task, alone and combined with the MLLR (maximum likelihood linear regression) adaptation.

A B S T R A C T This paper describes a new neural architecture for u i i-supervised learning of a classificat,ion of mixed t.rair-sient, signals. This method is I)asetl oii neural tkcli-niques for blind separation of sources and sul)space i-net,liotls. The feed-forward neural iiet,work dynaiii-ically builds and refreshes a n acoustic event.s classification by detecting novelties, creat,ing antl deleting classes. A self-organization process achieves a class prototype rotation in order to niinirnise the st,at,isti-cal dependence of class activities. Siiriulatd nirilt,i-tliniensional signals and rnixed acoustic signals i i i real noisy environment have been used to test, our inotlel. T h c result

This paper presents a neural architecture for model breaking detection in real world conditions. This technique use an Independent Component Classifier [1] for detection of unexpected or unknown events in noisy and varying environment. This method is based on subspace classifier [2] and Independant Component Analysis [3]. A feed-forward neural network adapts itself to input evolutions, by detecting novelties, creating and deleting classes. A second process achieves a prototype rotation in order to minimise mutual information of different classes. This synaptic weight evolution rule is based on an anti-hebbian learning rule inspired from neural methods for blind separation of sources [4]. Consequently, under the assumption of statistical independence of different classes, the system is able to detect novelties hidden by simultaneous acoustic events.

In the recent years, deep learning has become the leading approach to modern artificial intelligence (AI). The important improvement in terms of processing time required for learning AI based models alongside with the growing amount of available data made of deep neural networks (DNN) the strongest solution to solve complex real-world problems. However, a major challenge of artificial neural architectures lies on better considering the high-dimensionality of the data.To alleviate this issue, neural networks (NN) based on complex and hypercomplex algebras have been developped. The natural multidimensionality of the data is elegantly embedded within complex and hypercomplex neurons composing the model. In particular, quaternion neural networks (QNN) have been proposed to deal with up to four dimensional features, based on the quaternion representation of rotations and orientations. Unfortunately, and conversely to complex-valued neural networks that are nowadays known as a strong alternative to real-valued neural networks, QNNs suffer from numerous limitations that are carrefuly addressed in the different parts detailled in this thesis.The thesis consists in three parts that gradually introduce the missing concepts of QNNs, to make them a strong alternative to real-valued NNs. The first part introduces and list previous findings on quaternion numbers and quaternion neural networks to define the context and strong basics for building elaborated QNNs.The second part introduces state-of-the-art quaternion neural networks for a fair comparison with real-valued neural architectures. More precisely, QNNs were limited by their simple architectures that were mostly composed of a single and shallow hidden layer. In this part, we propose to bridge the gap between quaternion and real-valued models by presenting different quaternion architectures. First, basic paradigms such as autoencoders and deep fully-connected neural networks are introduced. Then, more elaborated convolutional and recurrent neural networks are extended to the quaternion domain. Experiments to compare QNNs over equivalents NNs have been conducted on real-world tasks across various domains, including computer vision, spoken language understanding and speech recognition. QNNs increase performances while reducing the needed number of neural parameters compared to real-valued neural networks.Then, QNNs are extended to unconventional settings. In a conventional QNN scenario, input features are manually segmented into three or four components, enabling further quaternion processing. Unfortunately, there is no evidence that such manual segmentation is the representation that suits the most to solve the considered task. Morevover, a manual segmentation drastically reduces the field of application of QNNs to four dimensional use-cases. Therefore the third part introduces a supervised and an unsupervised model to extract meaningful and disantengled quaternion input features, from any real-valued input signal, enabling the use of QNNs regardless of the dimensionality of the considered task. Conducted experiments on speech recognition and document classification show that the proposed approaches outperform traditional quaternion features., Au cours des dernières années, l’apprentissage profond est devenu l’approche privilégiée pour le développement d’une intelligence artificielle moderne (IA). L’augmentation importante de la puissance de calcul, ainsi que la quantité sans cesse croissante de données disponibles ont fait des réseaux de neurones profonds la solution la plus performante pour la resolution de problèmes complexes. Cependant, la capacité à parfaitement représenter la multidimensionalité des données réelles reste un défi majeur pour les architectures neuronales artificielles.Pour résoudre ce problème, les réseaux de neurones basés sur les algèbres des nombres complexes et hypercomplexes ont été développés. En particulier, les réseaux de neurones de quaternions (QNN) ont été proposés pour traiter les données tridi- mensionnelles et quadridimensionnelles, sur la base des quaternions représentant des rotations dans notre espace tridimensionnel. Malheureusement, et contrairement aux réseaux de neurones à valeurs complexes qui sont de nos jours acceptés comme une alternative aux réseaux de neurones réels, les QNNs souffrent de nombreuses lacunes qui sont en partie comblées par les différents travaux détaillés par ce manuscrit.Ainsi, la thèse se compose de trois parties qui introduisent progressivement les concepts manquants, afin de faire des QNNs une alternative aux réseaux neuronaux à valeurs réelles. La premiere partie présente et répertorie les précédentes découvertes relatives aux quaternions et aux réseaux de neurones de quaternions, afin de définir une base pour la construction des QNNs modernes.La deuxième partie introduit des réseaux neuronaux de quaternions état de l’art, afin de permettre une comparaison dans des contextes identiques avec les architectures modernes traditionnelles. Plus précisément, les QNNs étaient majoritairement limités par leurs architectures trop simples, souvent composées d’une seule couche cachée comportant peu de neurones. Premièrement, les paradigmes fondamentaux, tels que les autoencodeurs et les réseaux de neurones profonds sont présentés. Ensuite, les très répandus et étudiés réseaux de neurones convolutionnels et récurrents sont étendus à l’espace des quaternions. De nombreuses experiences sur différentes applications réelles, telles que la vision par ordinateur, la compréhension du langage parlé ainsi que la reconnaissance automatique de la parole sont menées pour comparer les modèles de quaternions introduits aux réseaux neuronaux conventionnels. Dans ces contextes bien spécifiques, les QNNs ont obtenus de meilleures performances ainsi qu’une réduction importante du nombre de paramètres neuronaux nécessaires à la phase d’apprentissage.Les QNNs sont ensuite étendus à des conditions d’entrainement permettant de traiter toutes les représentations en entrée des modèles de quaternions. Dans un scénario traditionnel impliquant des QNNs, les caractéristiques d’entrée sont manuellement segmentées en quatre composants, afin de correspondre à la representation induite par les quaternions. Malheureusement, il est difficile d’assurer qu’une telle segmentation est optimale pour résoudre le problème considéré. De plus, une segmentation manuelle réduit fondamentalement l’application des QNNs à des tâches naturellement définies dans un espace à au plus quatre dimensions. De ce fait, la troisième partie de cette thèse introduit un modèle supervisé et un modèle non supervisé permettant l’extraction de caractéristiques d’entrée désentrelacées et significatives dans l’espace des quaternions, à partir de n’importe quel type de signal réel uni-dimentionnel, permettant l’utilisation des QNNs indépendamment de la dimensionnalité des vecteurs d’entrée et de la tâche considérée. Les expériences menées sur la reconnaissance de la parole et la classification de documents parlés montrent que les approches proposées sont plus performantes que les représentations traditionnelles de quaternions.

In the same way as TV channels, data streams are represented as a sequence of successive events that can exhibit chronological relations (e.g. a series of programs, scenes, etc.). For a targeted channel, broadcast programming follows the rules defined by the channel itself, but can also be affected by the programming of competing ones. In such conditions, event sequences of parallel streams could provide additional knowledge about the events of a particular stream. In the sphere of machine learning, various methods that are suited for processing sequential data have been proposed. Long Short-Term Memory (LSTM) Recurrent Neural Networks have proven its worth in many applications dealing with this type of data. Nevertheless, these approaches are designed to handle only a single input sequence at a time. The main contribution of this thesis is about developing approaches that jointly process sequential data derived from multiple parallel streams. The application task of our work, carried out in collaboration with the computer science laboratory of Avignon (LIA) and the EDD company, seeks to predict the genre of a telecast. This prediction can be based on the histories of previous telecast genres in the same channel but also on those belonging to other parallel channels. We propose a telecast genre taxonomy adapted to such automatic processes as well as a dataset containing the parallel history sequences of 4 French TV channels. Two original methods are proposed in this work in order to take into account parallel stream sequences. The first one, namely the Parallel LSTM (PLSTM) architecture, is an extension of the LSTM model. PLSTM simultaneously processes each sequence in a separate recurrent layer and sums the outputs of each of these layers to produce the final output. The second approach, called MSE-SVM, takes advantage of both LSTM and Support Vector Machines (SVM) methods. Firstly, latent feature vectors are independently generated for each input stream, using the output event of the main one. These new representations are then merged and fed to an SVM algorithm. The PLSTM and MSE-SVM approaches proved their ability to integrate parallel sequences by outperforming, respectively, the LSTM and SVM models that only take into account the sequences of the main stream. The two proposed approaches take profit of the information contained in long sequences. However, they have difficulties to deal with short ones. Though MSE-SVM generally outperforms the PLSTM approach, the problem experienced with short sequences is more pronounced for MSE-SVM. Finally, we propose to extend this approach by feeding additional information related to each event in the input sequences (e.g. the weekday of a telecast). This extension, named AMSE-SVM, has a remarkably better behavior with short sequences without affecting the performance when processing long ones., Les flux de contenus audiovisuels peuvent être représentés sous forme de séquences d’événements (par exemple, des suites d’émissions, de scènes, etc.). Ces données séquentielles se caractérisent par des relations chronologiques pouvant exister entre les événements successifs. Dans le contexte d’une chaîne TV, la programmation des émissions suit une cohérence définie par cette même chaîne, mais peut également être influencée par les programmations des chaînes concurrentes. Dans de telles conditions,les séquences d’événements des flux parallèles pourraient ainsi fournir des connaissances supplémentaires sur les événements d’un flux considéré.La modélisation de séquences est un sujet classique qui a été largement étudié, notamment dans le domaine de l’apprentissage automatique. Les réseaux de neurones récurrents de type Long Short-Term Memory (LSTM) ont notamment fait leur preuve dans de nombreuses applications incluant le traitement de ce type de données. Néanmoins,ces approches sont conçues pour traiter uniquement une seule séquence d’entrée à la fois. Notre contribution dans le cadre de cette thèse consiste à élaborer des approches capables d’intégrer conjointement des données séquentielles provenant de plusieurs flux parallèles.Le contexte applicatif de ce travail de thèse, réalisé en collaboration avec le Laboratoire Informatique d’Avignon et l’entreprise EDD, consiste en une tâche de prédiction du genre d’une émission télévisée. Cette prédiction peut s’appuyer sur les historiques de genres des émissions précédentes de la même chaîne mais également sur les historiques appartenant à des chaînes parallèles. Nous proposons une taxonomie de genres adaptée à de tels traitements automatiques ainsi qu’un corpus de données contenant les historiques parallèles pour 4 chaînes françaises.Deux méthodes originales sont proposées dans ce manuscrit, permettant d’intégrer les séquences des flux parallèles. La première, à savoir, l’architecture des LSTM parallèles(PLSTM) consiste en une extension du modèle LSTM. Les PLSTM traitent simultanément chaque séquence dans une couche récurrente indépendante et somment les sorties de chacune de ces couches pour produire la sortie finale. Pour ce qui est de la seconde proposition, dénommée MSE-SVM, elle permet de tirer profit des avantages des méthodes LSTM et SVM. D’abord, des vecteurs de caractéristiques latentes sont générés indépendamment, pour chaque flux en entrée, en prenant en sortie l’événement à prédire dans le flux principal. Ces nouvelles représentations sont ensuite fusionnées et données en entrée à un algorithme SVM. Les approches PLSTM et MSE-SVM ont prouvé leur efficacité dans l’intégration des séquences parallèles en surpassant respectivement les modèles LSTM et SVM prenant uniquement en compte les séquences du flux principal. Les deux approches proposées parviennent bien à tirer profit des informations contenues dans les longues séquences. En revanche, elles ont des difficultés à traiter des séquences courtes.L’approche MSE-SVM atteint globalement de meilleures performances que celles obtenues par l’approche PLSTM. Cependant, le problème rencontré avec les séquences courtes est plus prononcé pour le cas de l’approche MSE-SVM. Nous proposons enfin d’étendre cette approche en permettant d’intégrer des informations supplémentaires sur les événements des séquences en entrée (par exemple, le jour de la semaine des émissions de l’historique). Cette extension, dénommée AMSE-SVM améliore remarquablement la performance pour les séquences courtes sans les baisser lorsque des séquences longues sont présentées.

The diachronic nature of broadcast news causes frequent variations in the linguisticcontent and vocabulary, leading to the problem of Out-Of-Vocabulary (OOV)words in automatic speech recognition. Most of the OOV words are found tobe proper names whereas proper names are important for automatic indexingof audio-video content as well as for obtaining reliable automatic transcriptions.New proper names missed by the speech recognition system can be recovered by adynamic vocabulary multi-pass recognition approach in which new proper namesare added to the speech recognition vocabulary based on the context of the spokencontent. Existing methods for vocabulary selection rely on web search engines andadaptation corpora and choose the new vocabulary words using term-documentfrequency and co-occurrence based features. Open vocabulary systems based onsub-word units are an interesting solution but they face the problem of producinga reliable text transcription. The goal of this thesis is to model the semantic andtopical context of new proper names in order to retrieve those which are relevantto the spoken content in the audio document. Training semantic/topic models is achallenging problem in this task because (a) several new proper names come witha low amount of data and (b) the context model should be robust to word errorsin the automatic transcription. Probabilistic topic models and word embeddingsfrom neural network models are explored for the task of retrieval of relevantproper names. A thorough evaluation of contextual representations from thesemodels is performed. It is argued that these representations, which are learnedin an unsupervised manner, are not the best for the given retrieval task. Neuralnetwork context models trained with an objective to maximise the retrieval performanceare proposed. A Neural Bag-of-Words (NBOW) model trained to learncontext vector representations at a document level is shown to outperform thegeneric representations. The proposed Neural Bag-of-Weighted-Words (NBOW2)model learns to assign a degree of importance to input words and has the abilityto capture task specic key-words. Experiments on automatic speech recognitionon French broadcast news videos demonstrate the eectiveness of the proposedmodels. Further evaluation of the NBOW2 model on standard text classicationtasks, including movie review sentiment classication and newsgroup topic classication, shows that it learns interesting information about the task and givesthe best classication accuracies among the bag-of-words models.; La nature diachronique des bulletins d'information provoque de fortes variationsdu contenu linguistique et du vocabulaire, conduisant au problème demots hors vocabulaire (Out-Of-Vocabulary OOV) dans le cadre de la reconnaissanceautomatique de la parole. La plupart des mots OOV se trouvent êtredes noms propres; ces noms propres sont importants pour l'indexation automatiquede ces contenus audio-vidéo, ainsi que pour obtenir des transcriptions fiables. Le but de cette these est de retrouver les noms propres manquants enutilisant le contexte an de les intégrer dans le lexique du système de reconnaissancepour effectuer une deuxièeme passe de reconnaissance. Les méthodesexistantes pour la sélection de nouveaux mots a ajouter au vocabulaire utilisentdes documents textuels récuperés sur le internet et choisissent les nouveaux motsde vocabulaire en utilisant des matrices terme-document ou des co-occurrencesde mots. Dans cette these nous allons modéliser le contexte sémantique etutiliser des informations thématiques contenus dans le document audio a transcrirean de récupérer les noms propres pertinents. Des modèles probabilistesde thème (topic model) et des projections dans un espace continu obtenues al'aide de réseaux de neurones (word embeddings) sont explorés pour la tâchede récuperation des noms propres pertinents. Une évaluation approfondie deces représentations contextuelles a été réalisée. Le contexte est estimé a l'aidede réseaux de neurones appris avec l'objectif de maximiser la récuperation desnoms propres pertinents. Un modèle de neurones (Neural Bag-of-Words, NBOW)entra^ne pour apprendre des représentations contextuelles au niveau du documentobtient de meilleures performances que les modèles classiques. Le nouveaumodèle que nous proposons (Bag-of-Weighted-Words, NBOW2) estime un degréd'importance pour chacun des mots du document et a la capacité de capturerdes mots importants spéciques. Des expériences de reconnaissance automatiquede bulletins d'information télévisés démontrent l'efficacité des modèles proposes.Une évaluation plus poussée du modèle NBOW2 sur les tâches de classicationde texte, d'analyse de critique de lm et de classication thématique de texteissus de groupe de discussion, démontre la capacité d'apprentissage du modèle etsurpasse les modèles utilisant des sac-de-mots (Bag-Of-Words, BOW).

The diachronic nature of broadcast news causes frequent variations in the linguisticcontent and vocabulary, leading to the problem of Out-Of-Vocabulary (OOV)words in automatic speech recognition. Most of the OOV words are found tobe proper names whereas proper names are important for automatic indexingof audio-video content as well as for obtaining reliable automatic transcriptions.New proper names missed by the speech recognition system can be recovered by adynamic vocabulary multi-pass recognition approach in which new proper namesare added to the speech recognition vocabulary based on the context of the spokencontent. Existing methods for vocabulary selection rely on web search engines andadaptation corpora and choose the new vocabulary words using term-documentfrequency and co-occurrence based features. Open vocabulary systems based onsub-word units are an interesting solution but they face the problem of producinga reliable text transcription. The goal of this thesis is to model the semantic andtopical context of new proper names in order to retrieve those which are relevantto the spoken content in the audio document. Training semantic/topic models is achallenging problem in this task because (a) several new proper names come witha low amount of data and (b) the context model should be robust to word errorsin the automatic transcription. Probabilistic topic models and word embeddingsfrom neural network models are explored for the task of retrieval of relevantproper names. A thorough evaluation of contextual representations from thesemodels is performed. It is argued that these representations, which are learnedin an unsupervised manner, are not the best for the given retrieval task. Neuralnetwork context models trained with an objective to maximise the retrieval performanceare proposed. A Neural Bag-of-Words (NBOW) model trained to learncontext vector representations at a document level is shown to outperform thegeneric representations. The proposed Neural Bag-of-Weighted-Words (NBOW2)model learns to assign a degree of importance to input words and has the abilityto capture task specic key-words. Experiments on automatic speech recognitionon French broadcast news videos demonstrate the eectiveness of the proposedmodels. Further evaluation of the NBOW2 model on standard text classicationtasks, including movie review sentiment classication and newsgroup topic classication, shows that it learns interesting information about the task and givesthe best classication accuracies among the bag-of-words models.; La nature diachronique des bulletins d'information provoque de fortes variationsdu contenu linguistique et du vocabulaire, conduisant au problème demots hors vocabulaire (Out-Of-Vocabulary OOV) dans le cadre de la reconnaissanceautomatique de la parole. La plupart des mots OOV se trouvent êtredes noms propres; ces noms propres sont importants pour l'indexation automatiquede ces contenus audio-vidéo, ainsi que pour obtenir des transcriptions fiables. Le but de cette these est de retrouver les noms propres manquants enutilisant le contexte an de les intégrer dans le lexique du système de reconnaissancepour effectuer une deuxièeme passe de reconnaissance. Les méthodesexistantes pour la sélection de nouveaux mots a ajouter au vocabulaire utilisentdes documents textuels récuperés sur le internet et choisissent les nouveaux motsde vocabulaire en utilisant des matrices terme-document ou des co-occurrencesde mots. Dans cette these nous allons modéliser le contexte sémantique etutiliser des informations thématiques contenus dans le document audio a transcrirean de récupérer les noms propres pertinents. Des modèles probabilistesde thème (topic model) et des projections dans un espace continu obtenues al'aide de réseaux de neurones (word embeddings) sont explorés pour la tâchede récuperation des noms propres pertinents. Une évaluation approfondie deces représentations contextuelles a été réalisée. Le contexte est estimé a l'aidede réseaux de neurones appris avec l'objectif de maximiser la récuperation desnoms propres pertinents. Un modèle de neurones (Neural Bag-of-Words, NBOW)entra^ne pour apprendre des représentations contextuelles au niveau du documentobtient de meilleures performances que les modèles classiques. Le nouveaumodèle que nous proposons (Bag-of-Weighted-Words, NBOW2) estime un degréd'importance pour chacun des mots du document et a la capacité de capturerdes mots importants spéciques. Des expériences de reconnaissance automatiquede bulletins d'information télévisés démontrent l'efficacité des modèles proposes.Une évaluation plus poussée du modèle NBOW2 sur les tâches de classicationde texte, d'analyse de critique de lm et de classication thématique de texteissus de groupe de discussion, démontre la capacité d'apprentissage du modèle etsurpasse les modèles utilisant des sac-de-mots (Bag-Of-Words, BOW).

These past ten years, TV series became increasingly popular. In contrast to classical TV series consisting of narratively self-sufficient episodes, modern TV series develop continuous plots over dozens of successive episodes. However, the narrative continuity of modern TV series directly conflicts with the usual viewing conditions: due to modern viewing technologies, the new seasons of TV series are being watched over short periods of time. As a result, viewers are largely disengaged from the plot, both cognitively and emotionally, when about to watch new seasons. Such a situation provides video summarization with remarkably realistic use-case scenarios, that we detail in Chapter 1. Furthermore, automatic movie summarization, long restricted to trailer generation based on low-level features, finds with TV series an unprecedented opportunity to address in well-defined conditions the so-called "semantic gap": summarization of narrative media requires content-oriented approaches capable to bridge the gap between low-level features and human understanding. We review in Chapter 2 the two main approaches adopted so far to address automatic movie summarization. Chapter 3 is dedicated to the various subtasks needed to build the intermediary representations on which our summarization framework relies: Section 3.1 focuses on video segmentation, whereas the rest of Chapter 3 is dedicated to the extraction of different mid-level features, either saliency-oriented (shot size, background music), or content-related (speakers). In Chapter 4, we make use of Social Network Analysis as a possible way to model the plot of modern TV series: the narrative dynamics can be properly captured by the evolution over time of the social network of interacting characters. Nonetheless, we have to address here the sequential nature of the narrative when taking instantaneous views of the state of the relationships between the characters. We show that standard time-windowing approaches can not properly handle this case, and we detail our own method for extracting dynamic social networks from narrative media. Chapter 5 is dedicated to the final generation and evaluation of character-oriented summaries, both able to reflect the plot dynamics and to emotionally re-engage viewers into the narrative. We evaluate our framework by performing a large-scale user study in realistic conditions.; Ces dix dernières années, les séries télévisées sont devenues de plus en plus populaires. Par opposition aux séries TV classiques composées d'épisodes auto-suffisants d'un point de vue narratif, les séries TV modernes développent des intrigues continues sur des dizaines d'épisodes successifs. Cependant, la continuité narrative des séries TV modernes entre directement en conflit avec les conditions usuelles de visionnage : en raison des technologies modernes de visionnage, les nouvelles saisons des séries TV sont regardées sur de courtes périodes de temps. Par conséquent, les spectateurs sur le point de visionner de nouvelles saisons sont largement désengagés de l'intrigue, à la fois d'un point de vue cognitif et affectif. Une telle situation fournit au résumé de vidéos des scénarios d'utilisation remarquablement réalistes, que nous détaillons dans le Chapitre 1. De plus, le résumé automatique de films, longtemps limité à la génération de bande-annonces à partir de descripteurs de bas niveau, trouve dans les séries TV une occasion inédite d'aborder dans des conditions bien définies ce qu'on appelle le "fossé sémantique" : le résumé de médias narratifs exige des approches orientées contenu, capables de jeter un pont entre des descripteurs de bas niveau et le niveau humain de compréhension. Nous passons en revue dans le Chapitre 2 les deux principales approches adoptées jusqu'ici pour aborder le problème du résumé automatique de films de fiction. Le Chapitre 3 est consacré aux différentes sous-tâches requises pour construire les représentations intermédiaires sur lesquelles repose notre système de génération de résumés : la Section 3.1 se concentre sur la segmentation de vidéos, tandis que le reste du chapitre est consacré à l'extraction de descripteurs de niveau intermédiaire, soit orientés saillance (échelle des plans, musique de fond), soit en relation avec le contenu (locuteurs). Dans le Chapitre 4, nous utilisons l'analyse des réseaux sociaux comme une manière possible de modéliser l'intrigue des séries TV modernes : la dynamique narrative peut être adéquatement capturée par l'évolution dans le temps du réseau des personnages en interaction. Cependant, nous devons faire face ici au caractère séquentiel de la narration lorsque nous prenons des vues instantanées de l'état des relations entre personnages. Nous montrons que les approches classiques par fenêtrage temporel ne peuvent pas traiter convenablement ce cas, et nous détaillons notre propre méthode pour extraire des réseaux sociaux dynamiques dans les médias narratifs. Le Chapitre 5 est consacré à la génération finale de résumés orientés personnages, capables à la fois de refléter la dynamique de l'intrigue et de ré-engager émotionnellement les spectateurs dans la narration. Nous évaluons notre système en menant à une large échelle et dans des conditions réalistes une enquête auprès d'utilisateurs.

In the Information Retrieval field, documents are usually considered as a "bagof-words". This model does not take into account the temporal structure of thedocument and is sensitive to noises which can alter its lexical form. These noisescan be produced by different sources : uncontrolled form of documents in microbloggingplatforms, automatic transcription of speech documents which are errorprone,lexical and grammatical variabilities in Web forums. . . The work presented inthis thesis addresses issues related to document representations from noisy sources.The thesis consists of three parts in which different representations of content areavailable. The first one compares a classical representation based on a term-frequencyrepresentation to a higher level representation based on a topic space. The abstractionof the document content allows us to limit the alteration of the noisy document byrepresenting its content with a set of high-level features. Our experiments confirm thatmapping a noisy document into a topic space allows us to improve the results obtainedduring different information retrieval tasks compared to a classical approach based onterm frequency. The major problem with such a high-level representation is that it isbased on a space theme whose parameters are chosen empirically.The second part presents a novel representation based on multiple topic spaces thatallow us to solve three main problems : the closeness of the subjects discussed in thedocument, the tricky choice of the "right" values of the topic space parameters and therobustness of the topic-based representation. Based on the idea that a single representationof the contents cannot capture all the relevant information, we propose to increasethe number of views on a single document. This multiplication of views generates "artificial"observations that contain fragments of useful information. The first experimentvalidated the multi-view approach to represent noisy texts. However, it has the disadvantageof being very large and redundant and of containing additional variability associatedwith the diversity of views. In the second step, we propose a method based onfactor analysis to compact the different views and to obtain a new robust representationof low dimension which contains only the informative part of the document whilethe noisy variabilities are compensated. During a dialogue classification task, the compressionprocess confirmed that this compact representation allows us to improve therobustness of noisy document representation.Nonetheless, during the learning process of topic spaces, the document is consideredas a "bag-of-words" while many studies have showed that the word position in a7document is useful. A representation which takes into account the temporal structureof the document based on hyper-complex numbers is proposed in the third part. Thisrepresentation is based on the hyper-complex numbers of dimension four named quaternions.Our experiments on a classification task have showed the effectiveness of theproposed approach compared to a conventional "bag-of-words" representation., En recherche d’information, les documents sont le plus souvent considérés comme des "sacs-de-mots". Ce modèle ne tient pas compte de la structure temporelle du document et est sensible aux bruits qui peuvent altérer la forme lexicale. Ces bruits peuvent être produits par différentes sources : forme peu contrôlée des messages des sites de micro-blogging, messages vocaux dont la transcription automatique contient des erreurs, variabilités lexicales et grammaticales dans les forums du Web. . . Le travail présenté dans cette thèse s’intéresse au problème de la représentation de documents issus de sources bruitées.La thèse comporte trois parties dans lesquelles différentes représentations des contenus sont proposées. La première partie compare une représentation classique utilisant la fréquence des mots à une représentation de haut-niveau s’appuyant sur un espace de thèmes. Cette abstraction du contenu permet de limiter l’altération de la forme de surface du document bruité en le représentant par un ensemble de caractéristiques de haut-niveau. Nos expériences confirment que cette projection dans un espace de thèmes permet d’améliorer les résultats obtenus sur diverses tâches de recherche d’information en comparaison d’une représentation plus classique utilisant la fréquence des mots.Le problème majeur d’une telle représentation est qu’elle est fondée sur un espace de thèmes dont les paramètres sont choisis empiriquement.La deuxième partie décrit une nouvelle représentation s’appuyant sur des espaces multiples et permettant de résoudre trois problèmes majeurs : la proximité des sujets traités dans le document, le choix difficile des paramètres du modèle de thèmes ainsi que la robustesse de la représentation. Partant de l’idée qu’une seule représentation des contenus ne peut pas capturer l’ensemble des informations utiles, nous proposons d’augmenter le nombre de vues sur un même document. Cette multiplication des vues permet de générer des observations "artificielles" qui contiennent des fragments de l’information utile. Une première expérience a validé cette approche multi-vues de la représentation de textes bruités. Elle a cependant l’inconvénient d’être très volumineuse,redondante, et de contenir une variabilité additionnelle liée à la diversité des vues. Dans un deuxième temps, nous proposons une méthode s’appuyant sur l’analyse factorielle pour fusionner les vues multiples et obtenir une nouvelle représentation robuste,de dimension réduite, ne contenant que la partie "utile" du document tout en réduisant les variabilités "parasites". Lors d’une tâche de catégorisation de conversations,ce processus de compression a confirmé qu’il permettait d’augmenter la robustesse de la représentation du document bruité.Cependant, lors de l’élaboration des espaces de thèmes, le document reste considéré comme un "sac-de-mots" alors que plusieurs études montrent que la position d’un terme au sein du document est importante. Une représentation tenant compte de cette structure temporelle du document est proposée dans la troisième partie. Cette représentation s’appuie sur les nombres hyper-complexes de dimension appelés quaternions. Nos expériences menées sur une tâche de catégorisation ont montré l’efficacité de cette méthode comparativement aux représentations classiques en "sacs-de-mots".

Robustness in speech recognition refers to the need to maintain high recognition accuracies even when the quality of the input speech is degraded. In the last decade, some papers proposed to use relevant meta-data in order to enhance the recognition process. Nevertheless, in many cases, an imperfect a priori transcript can be associated to the speech signal : movie subtitles, scenarios and theatrical plays, summariesand radio broadcast. This thesis addresses the issue of using such imperfect transcripts for improving the performance figures of automatic speech recognition (ASR) systems.Unfortunately, these a priori transcripts seldom correspond to the exact word utterances and suffer from a lack of temporal information. In spite of their varying quality, we will show how to use them to improve ASR systems. In the first part of the document we propose to integrate the imperfect transcripts inside the ASR search algorithm. We propose a method that allows us to drive an automatic speech recognition system by using prompts or subtitles. This driven decoding algorithm relies on an on-demand synchronization and on the linguistic rescoring of ASR hypotheses. In order to handle transcript excerpts, we suggest a method for extracting segments in large corpora. The second part presents the Driven Decoding Algorithm(DDA) approach in combining several speech recognition (ASR) systems : it consists in guiding the search algorithm of a primary ASR system by the one-best hypotheses of auxiliary systems.Our work suggests using auxiliary information directly inside an ASR system. The driven decoding algorithm enhances the baseline system and improves the a priori transcription. Moreover, the new combination schemes based on generalized-DDA significantly outperform state of the art combinations., L’utilisation des systèmes de reconnaissance automatique de la parole nécessite des conditions d’utilisation contraintes pour que ces derniers obtiennent des résultats convenables. Dans de nombreuses situations, des informations auxiliaires aux flux audio sont disponibles. Le travail de cette thèse s’articule autour des approches permettant d’exploiter ces transcriptions a priori disponibles. Ces informations se retrouvent dans de nombreuses situations : les pièces de théâtre avec les scripts des acteurs, les films accompagnés de sous-titres ou de leur scénario, les flashes d’information associés aux prompts des journalistes, les résumés d’émissions radio... Ces informations annexes sont de qualité variable, mais nous montrerons comment ces dernières peuvent être utilisées afin d’améliorer le décodage d’un SRAP.Ce document est divisé en deux axes liés par l’utilisation de transcriptions a priori au sein d’un SRAP : la première partie présente une méthode originale permettant d’exploiter des transcriptions a priori manuelles, et de les intégrer directement au cœur d’un SRAP. Nous proposons une méthode permettant de guider efficacement le système de reconnaissance à l’aide d’informations auxiliaires. Nous étendons notre stratégie à delarges corpus dénués d’informations temporelles. La seconde partie de nos travaux est axée sur la combinaison de SRAP. Nous proposons une combinaison de SRAP basée sur le décodage guidé : les transcriptions a priori guidant un SRAP principal sont fournies par des systèmes auxiliaires.Les travaux présentés proposent d’utiliser efficacement une information auxiliaire au sein d’un SRAP. Le décodage guidé par des transcriptions manuelles permet d’améliorer sensiblement la qualité du décodage ainsi que la qualité de la transcription a priori . Par ailleurs, les stratégies de combinaison proposées sont originales et obtiennent d’excellents résultats par rapport aux méthodes existantes à l’état de l’art.