Question answering over temporal knowledge graphs (KGs) efficiently uses facts contained in a temporal KG, which records entity relations and when they occur in time, to answer natural language questions (e.g., "Who was the president of the US before Obama?").These questions often involve three time-related challenges that previous work fail to adequately address: 1) questions often do not specify exact timestamps of interest (e.g., "Obama" instead of 2000); 2) subtle lexical differences in time relations (e.g., "before" vs "after"); 3) off-the-shelf temporal KG embeddings that previous work builds on ignore the temporal order of timestamps, which is crucial for answering temporal-order related questions.In this paper, we propose a time-sensitive question answering (TSQA) framework to tackle these problems.TSQA features a timestamp estimation module to infer the unwritten timestamp from the question.We also employ a time-sensitive KG encoder to inject ordering information into the temporal KG embeddings that TSQA is based on.With the help of techniques to reduce the search space for potential answers, TSQA significantly outperforms the previous state of the art on a new benchmark for question answering over temporal KGs, especially achieving a 32% (absolute) error reduction on complex questions that require multiple steps of reasoning over facts in the temporal KG.

In this study, we aim to predict the plausible future action steps given an observation of the past and study the task of instructional activity anticipation. Unlike previous anticipation tasks that aim at action label prediction, our work targets at generating natural language outputs that provide interpretable and accurate descriptions of future action steps. It is a challenging task due to the lack of semantic information extracted from the instructional videos. To overcome this challenge, we propose a novel knowledge distillation framework to exploit the related external textual knowledge to assist the visual anticipation task. However, previous knowledge distillation techniques generally transfer information within the same modality. To bridge the gap between the visual and text modalities during the distillation process, we devise a novel cross-modal contrastive distillation (CCD) scheme, which facilitates knowledge distillation between teacher and student in heterogeneous modalities with the proposed cross-modal distillation loss. We evaluate our method on the Tasty Videos dataset. CCD improves the anticipation performance of the visual-alone student model by a large margin of 40.2% relatively in BLEU4. Our approach also outperforms the state-of-the-art approaches by a large margin.

Document-level relation extraction (RE) poses new challenges compared to its sentence-level counterpart. One document commonly contains multiple entity pairs, and one entity pair occurs multiple times in the document associated with multiple possible relations. In this paper, we propose two novel techniques, adaptive thresholding and localized context pooling, to solve the multi-label and multi-entity problems. The adaptive thresholding replaces the global threshold for multi-label classification in the prior work with a learnable entities-dependent threshold. The localized context pooling directly transfers attention from pre-trained language models to locate relevant context that is useful to decide the relation. We experiment on three document-level RE benchmark datasets: DocRED, a recently released large-scale RE dataset, and two datasets CDRand GDA in the biomedical domain. Our ATLOP (Adaptive Thresholding and Localized cOntext Pooling) model achieves an F1 score of 63.4, and also significantly outperforms existing models on both CDR and GDA. We have released our code at https://github.com/wzhouad/ATLOP.

Self-attention mechanism in graph neural networks (GNNs) led to state-of-the-art performance on many graph representation learning tasks. Currently, at every layer, attention is computed between connected pairs of nodes and depends solely on the representation of the two nodes. However, such attention mechanism does not account for nodes that are not directly connected but provide important network context. Here we propose Multi-hop Attention Graph Neural Network (MAGNA), a principled way to incorporate multi-hop context information into every layer of attention computation. MAGNA diffuses the attention scores across the network, which increases the receptive field for every layer of the GNN. Unlike previous approaches, MAGNA uses a diffusion prior on attention values, to efficiently account for all paths between the pair of disconnected nodes. We demonstrate in theory and experiments that MAGNA captures large-scale structural information in every layer, and has a low-pass effect that eliminates noisy high-frequency information from graph data. Experimental results on node classification as well as the knowledge graph completion benchmarks show that MAGNA achieves state-of-the-art results: MAGNA achieves up to 5.7% relative error reduction over the previous state-of-the-art on Cora, Citeseer, and Pubmed. MAGNA also obtains the best performance on a large-scale Open Graph Benchmark dataset. On knowledge graph completion MAGNA advances state-of-the-art on WN18RR and FB15k-237 across four different performance metrics.

Commonsense knowledge graph (CKG) is a special type of knowledge graph (KG), where entities are composed of free-form text. Existing CKG completion methods focus on transductive learning setting, where all the entities are present during training. Here, we propose the first inductive learning setting for CKG completion, where unseen entities may appear at test time. We emphasize that the inductive learning setting is crucial for CKGs, because unseen entities are frequently introduced due to the fact that CKGs are dynamic and highly sparse. We propose InductivE as the first framework targeted at the inductive CKG completion task. InductivE first ensures the inductive learning capability by directly computing entity embeddings from raw entity attributes. Second, a graph neural network with novel densification process is proposed to further enhance unseen entity representation with neighboring structural information. Experimental results show that InductivE performs especially well on inductive scenarios where it achieves above 48% improvement over previous methods while also outperforms state-of-the-art baselines in transductive settings.

Xiaochen Hou, Peng Qi, Guangtao Wang, Rex Ying, Jing Huang, Xiaodong He, Bowen Zhou. Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2021.

Lingxiao Wang, Kevin Huang, Tengyu Ma, Quanquan Gu, Jing Huang. Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2021.

Recent years have seen great success in the use of neural seq2seq models on the text-to-SQL task. However, little work has paid attention to how these models generalize to realistic unseen data, which naturally raises a question: does this impressive performance signify a perfect generalization model, or are there still some limitations?In this paper, we first diagnose the bottleneck of the text-to-SQL task by providing a new testbed, in which we observe that existing models present poor generalization ability on rarely-seen data. The above analysis encourages us to design a simple but effective auxiliary task, which serves as a supportive model as well as a regularization term to the generation task to increase the models' generalization. Experimentally, We evaluate our models on a large text-to-SQL dataset WikiSQL. Compared to a strong baseline coarse-to-fine model, our models improve over the baseline by more than 3% absolute in accuracy on the whole dataset. More interestingly, on a zero-shot subset test of WikiSQL, our models achieve 5% absolute accuracy gain over the baseline, clearly demonstrating its superior generalizability.

Interpretable multi-hop reading comprehension (RC) over multiple documents is a challenging problem because it demands reasoning over multiple information sources and explaining the answer prediction by providing supporting evidences. In this paper, we propose an effective and interpretable Select, Answer and Explain (SAE) system to solve the multi-document RC problem. Our system first filters out answer-unrelated documents and thus reduce the amount of distraction information. This is achieved by a document classifier trained with a novel pairwise learning-to-rank loss. The selected answer-related documents are then input to a model to jointly predict the answer and supporting sentences. The model is optimized with a multi-task learning objective on both token level for answer prediction and sentence level for supporting sentences prediction, together with an attention-based interaction between these two tasks. Evaluated on HotpotQA, a challenging multi-hop RC data set, the proposed SAE system achieves top competitive performance in distractor setting compared to other existing systems on the leaderboard.

Distance-based knowledge graph embeddings have shown substantial improvement on the knowledge graph link prediction task, from TransE to the latest state-of-the-art RotatE.However, complex relations such as N-to-1, 1-to-N and N-to-N still remain challenging to predict.In this work, we propose a novel distance-based approach for knowledge graph link prediction.First we extend the RotatE from 2D complex domain to high dimensional space with orthogonal transforms to model relations.The orthogonal transform embedding for relations keeps the capability for modeling symmetric/anti-symmetric, inverse and compositional relations while achieves better modeling capacity.Second, the graph context is integrated into distance scoring functions directly.Specifically, graph context is explicitly modeled via two directed context representations.Each node embedding in knowledge graph is augmented with two context representations, which are computed from the neighboring outgoing and incoming nodes/edges respectively.The proposed approach improves prediction accuracy on the difficult N-to-1, 1-to-N and N-to-N cases.Our experimental results show that it achieves state-of-the-art results on two common benchmarks FB15k-237 and WNRR-18, especially on FB15k-237 which has many high in-degree nodes.Code available at https://github. com/JD-AI-Research-Silicon-Valley/ KGEmbedding-OTE.

Representation learning for speech emotion recognition is challenging due to labeled data sparsity issue and lack of gold-standard references. In addition, there is much variability from input speech signals, human subjective perception of the signals and emotion label ambiguity. In this paper, we propose a machine learning framework to obtain speech emotion representations by limiting the effect of speaker variability in the speech signals. Specifically we propose to disentangle the speaker characteristics from emotion through an adversarial training network in order to better represent emotion. Our method combines the gradient reversal technique with an entropy loss function to remove such speaker information. Our approach is evaluated on both IEMOCAP and CMU-MOSEI datasets. We show that our method improves speech emotion classification and increases generalization to unseen speakers.

The reliability of using fully convolutional networks (FCNs) has been successfully demonstrated by recent studies in many speech applications. One of the most popular variants of these FCNs is the `U-Net', which is an encoder-decoder network with skip connections. In this study, we propose `SkipConvNet' where we replace each skip connection with multiple convolutional modules to provide decoder with intuitive feature maps rather than encoder's output to improve the learning capacity of the network. We also propose the use of optimal smoothing of power spectral density (PSD) as a pre-processing step, which helps to further enhance the efficiency of the network. To evaluate our proposed system, we use the REVERB challenge corpus to assess the performance of various enhancement approaches under the same conditions. We focus solely on monitoring improvements in speech quality and their contribution to improving the efficiency of back-end speech systems, such as speech recognition and speaker verification, trained on only clean speech. Experimental findings show that the proposed system consistently outperforms other approaches.

Knowledge graph embedding has been an active research topic for knowledge base completion, with progressive improvement from the initial TransE, TransH, DistMult et al to the current state-of-the-art ConvE. ConvE uses 2D convolution over embeddings and multiple layers of nonlinear features to model knowledge graphs. The model can be efficiently trained and scalable to large knowledge graphs. However, there is no structure enforcement in the embedding space of ConvE. The recent graph convolutional network (GCN) provides another way of learning graph node embedding by successfully utilizing graph connectivity structure. In this work, we propose a novel end-to-end StructureAware Convolutional Network (SACN) that takes the benefit of GCN and ConvE together. SACN consists of an encoder of a weighted graph convolutional network (WGCN), and a decoder of a convolutional network called Conv-TransE. WGCN utilizes knowledge graph node structure, node attributes and edge relation types. It has learnable weights that adapt the amount of information from neighbors used in local aggregation, leading to more accurate embeddings of graph nodes. Node attributes in the graph are represented as additional nodes in the WGCN. The decoder Conv-TransE enables the state-of-the-art ConvE to be translational between entities and relations while keeps the same link prediction performance as ConvE. We demonstrate the effectiveness of the proposed SACN on standard FB15k-237 and WN18RR datasets, and it gives about 10% relative improvement over the state-of-theart ConvE in terms of HITS@1, HITS@3 and HITS@10.

Multi-hop reading comprehension (RC) across documents poses new challenge over single-document RC because it requires reasoning over multiple documents to reach the final answer. In this paper, we propose a new model to tackle the multi-hop RC problem. We introduce a heterogeneous graph with different types of nodes and edges, which is named as Heterogeneous Document-Entity (HDE) graph. The advantage of HDE graph is that it contains different granularity levels of information including candidates, documents and entities in specific document contexts. Our proposed model can do reasoning over the HDE graph with nodes representation initialized with co-attention and self-attention based context encoders. We employ Graph Neural Networks (GNN) based message passing algorithms to accumulate evidences on the proposed HDE graph. Evaluated on the blind test set of the Qangaroo WikiHop data set, our HDE graph based single model delivers competitive result, and the ensemble model achieves the state-of-the-art performance.

Machine reading comprehension (MRC) has attracted significant amounts of research attention recently, due to an increase of challenging reading comprehension datasets.In this paper, we aim to improve a MRC model's ability to determine whether a question has an answer in a given context (e.g. the recently proposed SQuAD 2.0 task).Our solution is a relation module that is adaptable to any MRC model.The relation module consists of both semantic extraction and relational information.We first extract high level semantics as objects from both question and context with multihead self-attentive pooling.These semantic objects are then passed to a relation network, which generates relationship scores for each object pair in a sentence.These scores are used to determine whether a question is nonanswerable.We test the relation module on the SQuAD 2.0 dataset using both the BiDAF and BERT models as baseline readers.We obtain 1.8% gain of F1 accuracy on top of the BiDAF reader, and 1.0% on top of the BERT base model.These results show the effectiveness of our relation module on MRC.

This paper aims to improve the widely used deep speaker embedding x-vector model. We propose the following improvements: (1) a hybrid neural network structure using both time delay neural network (TDNN) and long short-term memory neural networks (LSTM) to generate complementary speaker information at different levels; (2) a multi-level pooling strategy to collect speaker information from both TDNN and LSTM layers; (3) a regularization scheme on the speaker embedding extraction layer to make the extracted embeddings suitable for the following fusion step. The synergy of these improvements are shown on the NIST SRE 2016 eval test (with a 19% EER reduction) and SRE 2018 dev test (with a 9% EER reduction), as well as more than 10% DCF scores reduction on these two test sets over the x-vector baseline.

Speaker verification systems often degrade significantly when there is a language mismatch between training and testing data. Being able to improve cross-lingual speaker verification system using unlabeled data can greatly increase the robustness of the system and reduce human labeling costs. In this study, we introduce an unsupervised Adversarial Discriminative Domain Adaptation (ADDA) method to effectively learn an asymmetric mapping that adapts the target domain encoder to the source domain, where the target domain and source domain are speech data from different languages. ADDA, together with a popular Domain Adversarial Training (DAT) approach, are evaluated on a cross-lingual speaker verification task: the training data is in English from NIST SRE04-08, Mixer 6 and Switchboard, and the test data is in Chinese from AISHELL-I. We show that with the ADDA adaptation, Equal Error Rate (EER) of the x-vector system decreases from 9.331% to 7.645%, relatively 18.07% reduction of EER, and 6.32% reduction from DAT as well. Further data analysis of ADDA adapted speaker embedding shows that the learned speaker embeddings can perform well on speaker classification for the target domain data, and are less dependent with respect to the shift in language.

The I4U consortium was established to facilitate a joint entry to NIST speaker recognition evaluations (SRE). The latest edition of such joint submission was in SRE 2018, in which the I4U submission was among the best-performing systems. SRE'18 also marks the 10-year anniversary of I4U consortium into NIST SRE series of evaluation. The primary objective of the current paper is to summarize the results and lessons learned based on the twelve sub-systems and their fusion submitted to SRE'18. It is also our intention to present a shared view on the advancements, progresses, and major paradigm shifts that we have witnessed as an SRE participant in the past decade from SRE'08 to SRE'18. In this regard, we have seen, among others, a paradigm shift from supervector representation to deep speaker embedding, and a switch of research challenge from channel compensation to domain adaptation.

This work presents a novel approach for speaker diarization to leverage lexical information provided by automatic speech recognition. We propose a speaker diarization system that can incorporate word-level speaker turn probabilities with speaker embeddings into a speaker clustering process to improve the overall diarization accuracy. To integrate lexical and acoustic information in a comprehensive way during clustering, we introduce an adjacency matrix integration for spectral clustering. Since words and word boundary information for word-level speaker turn probability estimation are provided by a speech recognition system, our proposed method works without any human intervention for manual transcriptions. We show that the proposed method improves diarization performance on various evaluation datasets compared to the baseline diarization system using acoustic information only in speaker embeddings.

Recently, many e-commerce websites have encouraged their users to rate shopping items and write review texts. This review information has been very useful for understanding user preferences and item properties, as well as enhancing the capability to make personalized recommendations of these websites. In this paper, we propose to model user preferences and item properties using convolutional neural networks (CNNs) with dual local and global attention, motivated by the superiority of CNNs to extract complex features. By using aggregated review texts from a user and aggregated review text for an item, our model can learn the unique features (embedding) of each user and each item. These features are then used to predict ratings. We train these user and item networks jointly which enable the interaction between users and items in a similar way as matrix factorization. The local attention provides us insight on a user's preferences or an item's properties. The global attention helps CNNs focus on the semantic meaning of the whole review text. Thus, the combined local and global attentions enable an interpretable and better-learned representation of users and items. We validate the proposed models by testing on popular review datasets in Yelp and Amazon and compare the results with matrix factorization (MF), the hidden factor and topical (HFT) model, and the recently proposed convolutional matrix factorization (ConvMF+). Our proposed CNNs with dual attention model outperforms HFT and ConvMF+ in terms of mean square errors (MSE). In addition, we compare the user/item embeddings learned from these models for classification and recommendation. These results also confirm the superior quality of user/item embeddings learned from our model.

Deep belief networks (DBN) have shown impressive improvements over Gaussian mixture models for automatic speech recognition. In this work we use DBNs for audio-visual speech recognition; in particular, we use deep learning from audio and visual features for noise robust speech recognition. We test two methods for using DBNs in a multimodal setting: a conventional decision fusion method that combines scores from single-modality DBNs, and a novel feature fusion method that operates on mid-level features learned by the single-modality DBNs. On a continuously spoken digit recognition task, our experiments show that these methods can reduce word error rate by as much as 21% relative over a baseline multi-stream audio-visual GMM/HMM system.

Discriminatively trained full-covariance Gaussian mixture models have been shown to outperform its corresponding diagonal-covariance models on large vocabulary speech recognition tasks. However, the size of full-covariance model is much larger than that of diagonal-covariance model and is therefore not practical for use in a real system. In this paper, we present a method to build a large discriminatively trained full-covariance model with large (over 9000 hours) training corpora and still improve performance over the diagonal-covariance model. We then reduce the size of the full-covariance model to the size of its baseline diagonal-covariance model by using subspace constrained Gaussian mixture model (SCGMM). The resulting discriminatively trained SCGMM still retains the performance of its corresponding full-covariance model, and improves 5% relative over the same size diagonal-covariance model on a large vocabulary speech recognition task.

Modern speech applications utilize acoustic models with billions of parameters, and serve millions of users. Storing an acoustic model for each user is costly. We show through the use of sparse regularization, that it is possible to obtain competitive adaptation performance by changing only a small fraction of the parameters of an acoustic model. This allows for the compression of speaker-dependent models: a capability that has important implications for systems with millions of users. We achieve a performance comparable to the best Maximum A Posteriori (MAP) adaptation models while only adapting 5% of the acoustic model parameters. Thus it is possible to compress the speaker dependent acoustic models by close to a factor of 20. The proposed sparse adaptation criterion improves three aspects of previous work: It combines ℓ0and ℓ1penalties, have different adaptation rates for mean and variance parameters and is invariant to affine transformations.

Current hidden Markov acoustic modeling for large-vocabulary continuous speech recognition (LVCSR) heavily relies on the availability of abundant labeled transcriptions. Given that speech labeling is both expensive and time-consuming while there is a huge amount of unlabeled data easily available nowadays, the semi-supervised learning (SSL) from both labeled and unlabeled data aiming to reduce the development cost for LVCSR becomes more important than ever. In this paper, a new SSL approach is proposed which exploits the cross-view transfer learning for LVCSR through a committee machine consisting of multiple views learned from different acoustic features and randomized decision trees. In addition, a multi-objective learning scheme is developed in each view by maximizing a hybrid information-theoretic criterion which is established by the relative entropy between labeled data and their labels and the entropy of unlabeled data. The multi-objective scheme is then generalized to a unified SSL framework which can be interpreted into a variety of learning strategies under different weighting schemes. Experiments conducted on English Broadcast News using 50 hours of transcribed speech with 50 hours and 150 hours of untranscribed speech show the benefits of proposed approaches.

Maximum A Posteriori (MAP) adaptation is a powerful tool for building speaker specific acoustic models. Modern speech applications utilize acoustic models with millions of parameters, and serve millions of users. Storing an acoustic model for each user in such settings is costly. However, speaker specific acoustic models are generally similar to the acoustic model being adapted. By imposing sparseness constraints, we can save significantly on storage, and even improve the quality of the resulting speaker-dependent model. In this paper we utilize the ℓ1or ℓ0norm as a regularizer to induce sparsity. We show that we can obtain up to 95% sparsity with negligible loss in recognition accuracy, with both penalties. By removing small differences, which constitute “adaptation noise”, sparse MAP is actually able to improve upon MAP adaptation. Sparse MAP reduces the MAP word error rate by 2% relative at 89% sparsity.

Current hidden Markov acoustic modeling for large vocabulary continuous speech recognition (LVCSR) relies on the availability of abundant labeled transcriptions. Given that speech labeling is both expensive and time-consuming while there is a huge amount of unlabeled data easily available nowadays, semi-supervised learning (SSL) from both labeled and unlabeled data which aims to reduce the development cost for LVCSR becomes more important than ever. In this paper, we propose SSL for LVCSR by using the multiple views learned from different acoustic features and randomized decision trees. In addition, we develop the multi-objective learning of HMM-based acoustic models by optimizing a hybrid criterion which is established by the combination of the discriminative mutual information from labeled data and the entropy from unlabeled data. Experiments conducted on Broadcast News show the benefits of proposed methods.

Feature-space transforms such as feature-space maximum likelihood linear regression (FMLLR) are very effective speaker adaptation technique, especially on mismatched test data. In this study, we extend the full-rank square matrix of FMLLR to a non-square matrix that uses neighboring feature vectors in estimating the adapted central feature vector. Through optimizing an appropriate objective function we aim to filter out and transform features through the correlation of the feature context. We compare to FMLLR that just con sider the current feature vector only. Our experiments are conducted on the automobile data with different speed conditions. Results show that context filtering improves 23% on word error rate over conventional FMLLR on noisy 60mph data with adapted ML model, and 7%/9% improvement over the discriminatively trained FMMI/BMMI models.

HMM-based audio-visual speech recognition (AVSR) systems have shown success in continuous speech recognition by combining visual and audio information, especially in noisy environments. In this paper we study how to improve decision trees used to create context classes in HMM-based AVSR systems. Traditionally, visual models have been trained with the same context classes as the audio only models. In this paper we investigate the use of separate decision trees to model the context classes for the audio and visual streams independently. Additionally we investigate the use of viseme classes in the decision tree building for the visual stream. On experiments with a 37-speaker 1.5 hours test set (about 12000 words) of continuous digits in noise, we obtain about a 3% absolute (20% relative) gain on AVSR performance by using separate decision trees for the audio and visual streams when using viseme classes in decision tree building for the visual stream.

Smart homes for the aging population have recently started attracting the attention of the research community. One of the problems of interest is this of monitoring the activities of daily living (ADLs) of the elderly, in order to help identify critical problems, aiming to improve their protection and general well-being. In this paper, we report on our initial attempts to recognize such activities, based on input from networks of far-field microphones distributed inside the home. We propose two approaches to the problem: The first models the entire activity, which typically covers long time spans, with a single statistical model, for example a hidden Markov model (HMM), a Gaussian mixture model (GMM), or GMM super-vectors in conjunction with support vector machines (SVMs). The second is a two-step approach: It first performs acoustic event detection (AED) to locate distinctive events, characteristic of the ADLs, and it is subsequently followed by a post-processing stage that employs activity-specific language models (LMs) to classify the output sequences of detected events into ADLs. Experiments are reported on a corpus containing a small number of acted ADLs, collected as part of the Netcarity Integrated Project inside a two-room smart home. Our results show that SVM GMM supervector modeling improves six-class ADL classification accuracy to 76%, compared to 56% achieved by the GMMs, while also outperforming HMMs by 8% absolute. Preliminary results from LM scoring of acoustic event sequences are comparable to those from GMMs on a three-class ADL classification task.

We present a system that detects human falls in the home environment, distinguishing them from competing noise, by using only the audio signal from a single far-field microphone. The proposed system models each fall or noise segment by means of a Gaussian mixture model (GMM) supervector, whose Euclidean distance measures the pairwise difference between audio segments. A support vector machine built on a kernel between GMM supervectors is employed to classify audio segments into falls and various types of noise. Experiments on a dataset of human falls, collected as part of the Netcarity project, show that the method improves fall classification F-score to 67% from 59% of a baseline GMM classifier. The approach also effectively addresses the more difficult fall detection problem, where audio segment boundaries are unknown. Specifically, we employ it to reclassify confusable segments produced by a dynamic programming scheme based on traditional GMMs. Such post-processing improves a fall detection accuracy metric by 5% relative.

Speaker diarization is often performed before automatic speech recognition (ASR) to label speaker segments. In this paper we present two simple schemes to improve the speaker diarization performance. The first is to iteratively refine GMM speaker models by frame level re-labelingand smoothingof the decision likelihood. The second is to use word level alignment information from the ASR process. We focus on the CHIL lecture meeting data. Our experiments on the NIST RT06 evaluation data show that these simple methods are quite effective in improving our baseline diarization system, with alignment information providing 1% absolute reduction in diarization error rate (DER) and the re-label smoothing providing an additional 3.51% absolute reduction in DER. The overall system generates a DER that is 6.8% relative better than the top performing system from the RT06 evaluation.

Speech processing of lectures recorded inside smart rooms has recently attracted much interest. In particular, the topic has been central to the Rich Transcription (RT) Meeting Recognition Evaluation campaign series, sponsored by NIST, with emphasis placed on benchmarking speech activity detection (SAD), speaker diarization (SPKR), speech-to-text (STT), and speakerattributed STT (SASTT) technologies. In this paper, we present the IBM systems developed to address these tasks in preparation for the RT 2007 evaluation, focusing on the far-field condition of lecture data collected as part of European project CHIL. For their development, the systems are benchmarked on a subset of the RT Spring 2006 (RT06s) evaluation test set, where they yield significant improvements for all SAD, SPKR, and STT tasks over RT06s results; for example, a 16% relative reduction in word error rate is reported in STT, attributed to a number of system advances discussed here. Initial results are also presented on SASTT, a task newly introduced in 2007 in place of the discontinued SAD. Index Terms: speech processing, speech recognition, speaker diarization, speech activity detection, lectures, smart rooms.

Multi-stream hidden Markov models (HMMs) have recently been very successful in audio-visual speech recognition, where the audio and visual streams are fused at the final decision level. In this paper we investigate fast feature space speaker adaptation using multi-stream HMMs for audio-visual speech recognition. In particular, we focus on studying the performance of feature-space maximum likelihood linear regression (fMLLR), a fast and effective method for estimating feature space transforms. Unlike the common speaker adaptation techniques of MAP or MLLR, fMLLR does not change the audio or visual HMM parameters, but simply applies a single transform to the testing features. We also address the problem of fast and robust on-line fMLLR adaptation using feature space maximum a posterior linear regression (fMAPLR). Adaptation experiments are reported on the IBM infrared headset audio-visual database. On average for a 20-speaker 1 hour independent test set, the multi-stream fMLLR achieves 31% relative gain on the clean audio condition, and 59% relative gain on the noisy audio condition (approximately 7 dB) as compared to the baseline multi-stream system

fMPE is a recently introduced discriminative training technique that uses the Minimum Phone Error (MPE) discriminative criterion to train a feature-level transformation. In this paper we investigate fMPE trained audio/visual features for multistream HMM-based audio-visual speech recognition. A flexible, layer-based implementation of fMPE allows us to combine the the visual information with the audio stream using the discriminative traning process, and dispense with the multiple stream approach. Experiments are reported on the IBM infrared headset audio-visual database. On average of 20-speaker 1 hour speaker independent test data, the fMPE trained acoustic features achieve 33% relative gain. Adding video layers on top of audio layers gives additional 10% gain over fMPE trained features from the audio stream alone. The fMPE trained visual features achieve 14% relative gain, while the decision fusion of audio/visual streams with fMPE trained features achieves 29% relative gain. However, fMPE trained models do not improve over the original models on the mismatched noisy test data.

Much progress has been achieved during the past two decades in audio-visual automatic speech recognition (AVASR). However, challenges persist that hinder AVASR deployment in practical situations, most notably, robust and fast extraction of visual speech features. We review our efforts in overcoming this problem, based on an appearance-based visual feature representation of the speaker's mouth region. We cover three topics in particular. Firstly, we discuss AVASR in realistic, visually challenging domains, where lighting, background, and head-pose vary significantly. To enhance visual-front-end robustness in such environments, we employ an improved statistical-based face detection algorithm that significantly outperforms our baseline scheme. However, visual-only recognition remains inferior to visually "clean" (studio-like) data, thus demonstrating the importance of accurate mouth region extraction. We then consider a wearable audio-visual sensor to capture the mouth region directly, thus eliminating face detection. Its use improves visual-only recognition, even over full-face videos recorded in the studio-like environment. Finally, we address the speed issue in visual feature extraction, by discussing our real-time AVASR prototype implementation. The reported progress demonstrates the feasibility of practical AVASR.

Digital archives have emerged as the pre-eminent method for capturing the human experience. Before such archives can be used efficiently, their contents must be described. The NSF-funded MALACH project aims to provide improved access to large spoken archives by advancing the state-of-the-art in automated speech recognition (ASR), Information Retrieval (IR) and related technologies [1,2] for multiple languages. This paper describes the ASR research for the English speech in the MALACH corpus. The MALACH corpus consists of unconstrained, natural speech filled with disfluencies, heavy accents, age-related coarticulation, uncued speaker and language switching, and emotional speech collected in the form of interviews from over 52000 speakers in 32 languages. In this paper, we describe this new testbed for developing speech recognition algorithms and report on the performance of well-known techniques for building better acoustic models for the speaking styles seen in this corpus. The best English ASR system to date has a word error rate of 43.8% on this corpus.

This paper addresses the influence of audio segmentation and segment clustering on automatic transcription accuracy for large spoken archives. The work forms part of the ongoing MALACH project, which is developing advanced techniques for supporting access to the world’s largest digital archive of video oral histories collected in many languages from over 52000 survivors and witnesses of the Holocaust. We present several audio-only and audio-visual segmentation schemes, including two novel schemes: the first is iterative and audio-only, the second uses audio-visual synchrony. Unlike most previous work, we evaluate these schemes in terms of their impact upon recognition accuracy. Results on English interviews show the automatic segmentation schemes give performance comparable to (exhorbitantly expensive and impractically lengthy) manual segmentation when using a single pass decoding strategy based on speaker-independent models. However, when using a multiple pass decoding strategy with adaptation, results are sensitive to both initial audio segmentation and the scheme for clustering segments prior to adaptation: the combination of our best automatic segmentation and clustering scheme has an error rate 8% worse (relative) to manual audio segmentation and clustering due to the occurrence of “speaker-impure ” segments. 1.

This paper applies the recently proposed Extended Maximum Likelihood Linear Transformation (EMLLT) model in a Speaker Adaptive Training (SAT) context on the Switchboard database. Adaptation is carried out with maximum likelihood estimation of linear transforms for the means, precisions (inverse covariances) and the feature-space under the EMLLT model. This paper shows the first experimental evidence that significant word-error-rate improvements can be achieved with the EMLLT model (in both VTL and VTL+SAT training contexts) over a state-of-the-art diagonal covariance model in a difficult large-vocabulary conversational speech recognition task. The improvements were of the order of 1% absolute in multiple scenarios.

We report on the performance of automatic speech recognition (ASR) systems on voicemail transcription. Voicemail is spontaneous telephone speech recorded over a variety of channels; consequently, it is representative of many challenging problems in speech recognition. In the course of working on this task, several algorithms were developed that focus on different components of an ASR system, including lexicon design, feature extraction, hypothesis search, and adaptation. We report the improvements provided by these techniques, as well as other standard techniques, on a voicemail test set. Although the techniques are benchmarked on voicemail test data, their scope is not restricted to this domain as they address fundamental aspects of the speech recognition process.

In this paper we address the problem of extracting key pieces of information from voicemail messages, such as the identity and phone number of the caller. This task differs from the named entity task in that the information we are interested in is a subset of the named entities in the message, and consequently, the need to pick the correct subset makes the problem more difficult. Also, the caller's identity may include information that is not typically associated with a named entity. In this work, we present three information extraction methods, one based on hand-crafted rules, one based on maximum entropy tagging, and one based on probabilistic transducer induction. We evaluate their performance on both manually transcribed messages and on the output of a speech recognition system.

In this paper we report recent improvements in word error performance on a voicemail transcription task. Last year, the speaker independent word error rate (WER) on the dev test set of the Voicemail Transcription task was reported at 35.45% [1]. This year, we report a relative 20% gain over this number. The improvements were obtained using several new algorithms and an increased amount of training data. In addition to benchmarking the performance of these algorithms on the Voicemail task, we have also evaluated them on the Switchboard task, and we report these results here as well. Finally, we also present the result of crossdomain experiments to evaluate the domain-independence of the constructed systems. 1. INTRODUCTION In this paper we report recent improvements in transcribing conversational telephone speech, as typified by the Voicemail and Switchboard transcription tasks. These improvements are a result of some new algorithms and, in the case of Voicemail, also due to an increa...

Speaker adaptation techniques have emerged as very effective and practical methods to improve ASR performance on a test speaker with only limited speech data from the speaker. We explore the use of adaptation techniques on a new Voicemail database and present some theoretical extensions of the Cluster Transformation (CT) technique. Our experiments on 40 hours of voicemail data and four clusters shows that using cluster information with MLLR improves over baseline MLLR by 2.2% (relative). When the amount of adaptation data in a short message is insufficient to reliably decide its cluster, higher improvements result when we use MLLR for the very short messages and CT on longer ones. 1. INTRODUCTION Speaker adaptation techniques have emerged as very effective and practical methods to improve ASR performance on a test speaker with the use of a limited amount of speech data from the speaker. Most of these techniques generally start with speaker-independent (SI) acoustic models, and adapt ...

In this paper we report recent improvements in voicemail transcription. Last year, the speaker independent and speaker adapted word error rates (WER) on the Voicemail Transcription task were reported at 41.94% and 38.18% respectively. This year, we report a relative improvement of 18% in the speaker independent performance and 11% in the speaker adapted performance over last year. This improvement is a result of some new algorithms and an increase in the amount of training data. In the following sections, we describe the contribution of several components to improving the word error rate. 1. INTRODUCTION In this paper we report recent improvements in voicemail transcription. The voicemail transcription task was introduced last year [1] as representing a style of conversational telephone speech that is somewhat different from the Switchboard and CallHome databases. Last year, the speaker independent and speaker adapted word error rates (WER) on this task were reported at 41.94% and 38...

We suggest the use of the color correlogram as a generic indexing tool to tackle various computer vision problems. Correlograms were shown to be very effective for content-based image retrieval. We adapt the correlogram to handle the problems of image subregion querying, object localization, object tracking, and cut detection. Experimental results suggest that the color correlogram is much more effective than the histogram for these applications, with insignificant additional computational, storage, or processing cost. We also provide a technique to cut down the storage requirement of correlograms so that it is the same as that of histograms, with only negligible performance penalty compared to the original correlogram.

Organizing images into semantic categories can be extremely useful for searching and browsing through large collections of images. Not much work has been done on automatic image classification, however. In this paper, we propose a method for hierarchical classification of images via supervised learning. This scheme relies on using a good low-level feature and subsequently performing feature-space reconfiguration using singular value decomposition to reduce noise and dimensionality. We use the training data to obtain a hierarchical classification tree that can be used to categorize new images. Our experimental results suggest that this scheme not only performs better than standard nearest-neighbor techniques, but also has both storage and computational advantages. 1 Introduction The proliferation of the world-wide web has given easy access to an explosively growing volume of visual data. Unfortunately, this data on the web is both scattered and unorganized, making search and retrieval...

We define a new image feature called the color correlogram and use it for image indexing and comparison. This feature distills the spatial correlation of colors, and is both effective and inexpensive for content-based image retrieval. The correlogram robustly tolerates large changes in appearance and shape caused by changes in viewing positions, camera zooms, etc. Experimental evidence suggests that this new feature outperforms not only the traditional color histogram method but also the recently proposed histogram refinement methods for image indexing/retrieval.

The paper addresses how relevance feedback can be used to improve the performance of content-based image retrieval.We present two supervised learning methods: learning the query and learning the metric.We combine the leaming methods with the recently proposed color correlograms for image indexing/retrieval.Our results on a large image database of over 20,000 images suggest that these learning methods are quite effective for content-based image retrieval.