- Date & Time: Tuesday, May 7, 2013; 2:30 PM
Speaker: Dr. Yotaro Kubo, NTT Communication Science Laboratories, Kyoto, Japan
Research Area: Speech & Audio
Abstract 
Kernel methods are important to realize both convexity in estimation and ability to represent nonlinear classification. However, in automatic speech recognition fields, kernel methods are not widely used conventionally. In this presentation, I will introduce several attempts to practically incorporate kernel methods into acoustic models for automatic speech recognition. The presentation will consist of two parts. The first part will describes maximum entropy discrimination and its application to a kernel machine training. The second part will describes dimensionality reduction of kernel-based features.
-
- Date: May 2, 2013
Where: International Conference on Learning Representations (ICLR)
MERL Contact: Jonathan Le Roux
Research Area: Speech & Audio
Brief - The paper "Block Coordinate Descent for Sparse NMF" by Potluru, V.K., Plis, S.M., Le Roux, J., Pearlmutter, B.A., Calhoun, V.D. and Hayes, T.P. was presented at the International Conference on Learning Representations (ICLR).
-
- Date: March 1, 2013
Where: IEEE Signal Processing Letters
MERL Contact: Jonathan Le Roux
Research Area: Speech & Audio
Brief - The article "Consistent Wiener Filtering for Audio Source Separation" by Le Roux, J. and Vincent, E. was published in IEEE Signal Processing Letters.
-
- Date & Time: Tuesday, February 26, 2013; 12:00 PM
Speaker: Prof. Taylan Cemgil, Bogazici University, Istanbul, Turkey
MERL Host: Jonathan Le Roux
Research Area: Speech & Audio
Abstract 
Algorithms for decompositions of matrices are of central importance in machine learning, signal processing and information retrieval, with SVD and NMF (Nonnegative Matrix Factorisation) being the most widely used examples. Probabilistic interpretations of matrix factorisation models are also well known and are useful in many applications (Salakhutdinov and Mnih 2008; Cemgil 2009; Fevotte et. al. 2009). In the recent years, decompositions of multiway arrays, known as tensor factorisations have gained significant popularity for the analysis of large data sets with more than two entities (Kolda and Bader, 2009; Cichocki et. al. 2008). We will discuss a subset of these models from a statistical modelling perspective, building upon probabilistic Bayesian generative models and generalised linear models (McCulloch and Nelder). In both views, the factorisation is implicit in a well-defined hierarchical statistical model and factorisations can be computed via maximum likelihood.
We express a tensor factorisation model using a factor graph and the factor tensors are optimised iteratively. In each iteration, the update equation can be implemented by a message passing algorithm, reminiscent to variable elimination in a discrete graphical model. This setting provides a structured and efficient approach that enables very easy development of application specific custom models, as well as algorithms for the so called coupled (collective) factorisations where an arbitrary set of tensors are factorised simultaneously with shared factors. Extensions to full Bayesian inference for model selection, via variational approximations or MCMC are also feasible. Well known models of multiway analysis such as Nonnegative Matrix Factorisation (NMF), Parafac, Tucker, and audio processing (Convolutive NMF, NMF2D, SF-SSNTF) appear as special cases and new extensions can easily be developed. We will illustrate the approach with applications in link prediction and audio and music processing.
-
- Date & Time: Monday, January 28, 2013; 11:00 AM
Speaker: Prof. Jen-Tzung Chien, National Chiao Tung University, Taiwan
Research Area: Speech & Audio
Abstract 
Bayesian learning provides attractive tools to model, analyze, search, recognize and understand real-world data. In this talk, I will introduce a new Bayesian group sparse learning and its application on speech recognition and signal separation. First of all, I present the group sparse hidden Markov models (GS-HMMs) where a sequence of acoustic features is driven by Markov chain and each feature vector is represented by two groups of basis vectors. The features across states and within states are represented accordingly. The sparse prior is imposed by introducing the Laplacian scale mixture (LSM) distribution. The robustness of speech recognition is illustrated. On the other hand, the LSM distribution is also incorporated into Bayesian group sparse learning based on the nonnegative matrix factorization (NMF). This approach is developed to estimate the reconstructed rhythmic and harmonic music signals from single-channel source signal. The Monte Carlo procedure is presented to infer two groups of parameters. The future work of Bayesian learning shall be discussed.
-
- Date & Time: Tuesday, December 11, 2012; 12:00 PM
Speaker: Takahiro Oku, NHK Science & Technology Research Laboratories
Research Area: Speech & Audio
Abstract - In this talk, I will present human-friendly broadcasting research conducted in NHK and research on speech recognition for real-time closed-captioning. The goal of human-friendly broadcasting research is to make broadcasting more accessible and enjoyable for everyone, including children, elderly, and physically challenged persons. The automatic speech recognition technology that NHK has developed makes it possible to create captions for the hearing impaired in real-time automatically. For sports programs such as professional sumo wrestling, a closed-captioning system has already been implemented in which captions are created by using speech recognition on a captioning re-speaker. In 2011, NHK General Television started broadcasting of closed captions for the information program "Morning Market". After the introduction of the implemented closed-captioning system, I will talk about our recent improvement obtained by an adaptation method that creates a more effective acoustic model using error correction results. The method reflects recognition error tendencies more effectively.
-
- Date: December 6, 2012
Where: APSIPA Transactions on Signal and Information Processing
Research Area: Speech & Audio
Brief - The article "Bayesian Approaches to Acoustic Modeling: A Review" by Watanabe, S. and Nakamura, A. was published in APSIPA Transactions on Signal and Information Processing.
-
- Date: November 28, 2012
Where: Techniques for Noise Robustness in Automatic Speech Recognition
MERL Contact: Jonathan Le Roux
Research Area: Speech & Audio
Brief - The article "Factorial Models for Noise Robust Speech Recognition" by Hershey, J.R., Rennie, S.J. and Le Roux, J. was published in the book Techniques for Noise Robustness in Automatic Speech Recognition.
-
- Date: November 1, 2012
Where: IEEE Signal Processing Magazine
Research Area: Speech & Audio
Brief - The article "Structured Discriminative Models For Speech Recognition" by Gales, M., Watanabe, S. and Fosler-Lussier, E. was published in IEEE Signal Processing Magazine.
-
- Date & Time: Wednesday, October 24, 2012; 3:20 PM
Speaker: Dr. Steven J. Rennie, IBM Research
MERL Host: Jonathan Le Roux
Research Area: Speech & Audio
-
- Date & Time: Wednesday, October 24, 2012; 9:10 AM
Speaker: Prof. Jim Glass and Chia-ying Lee, MIT CSAIL
MERL Host: Jonathan Le Roux
Research Area: Speech & Audio
-
- Date & Time: Wednesday, October 24, 2012; 4:05 PM
Speaker: Dr. John R. Hershey, MERL
MERL Host: Jonathan Le Roux
Research Area: Speech & Audio
-
- Date & Time: Wednesday, October 24, 2012; 11:00 AM
Speaker: Prof. Dan Ellis, Columbia University
MERL Host: Jonathan Le Roux
Research Area: Speech & Audio
-
- Date & Time: Wednesday, October 24, 2012; 8:30 AM - 5:00 PM
Location: MERL
MERL Contact: Jonathan Le Roux
Research Area: Speech & Audio
Brief - SANE 2012, a one-day event gathering researchers and students in speech and audio from the northeast of the American continent, will be held on Wednesday October 24, 2012 at Mitsubishi Electric Research Laboratories (MERL) in Cambridge, MA.
-
- Date & Time: Wednesday, October 24, 2012; 9:55 AM
Speaker: Dr. Tara Sainath, IBM Research
MERL Host: Jonathan Le Roux
Research Area: Speech & Audio
-
- Date & Time: Wednesday, October 24, 2012; 2:15 PM
Speaker: Dr. Herb Gish, BBN - Raytheon
MERL Host: Jonathan Le Roux
Research Area: Speech & Audio
-
- Date & Time: Wednesday, October 24, 2012; 11:45 AM
Speaker: Josh McDermott, MIT, BCS
MERL Host: Jonathan Le Roux
Research Area: Speech & Audio
-
- Date & Time: Wednesday, October 24, 2012; 1:30 PM
Speaker: Dr. Timothy J. Hazen and David Harwath, MIT Lincoln Labs / MIT CSAIL
MERL Host: Jonathan Le Roux
Research Area: Speech & Audio
-
- Date: October 22, 2012
Where: Annual Meeting of the Human Factors and Ergonomics Society (HFES)
Research Area: Speech & Audio
Brief - The paper "Evaluation of Two Types of In-Vehicle Music Retrieval and Navigation Systems" by Zhang, J., Borowsky, A., Schmidt-Nielsen, B., Harsham, B., Weinberg, G., Romoser, M.R.E. and Fisher, D.L. was presented at the Annual Meeting of the Human Factors and Ergonomics Society (HFES).
-
- Date & Time: Thursday, October 11, 2012; 2:30 PM
Speaker: Dr. Gautham J. Mysore, Adobe
Research Area: Speech & Audio
Abstract - Non-negative spectrogram factorization techniques have become quite popular in the last decade as they are effective in modeling the spectral structure of audio. They have been extensively used for applications such as source separation and denoising. These techniques however fail to account for non-stationarity and temporal dynamics, which are two important properties of audio. In this talk, I will introduce the non-negative hidden Markov model (N-HMM) and the non-negative factorial hidden Markov model (N-FHMM) to model single sound sources and sound mixtures respectively. They jointly model the spectral structure and temporal dynamics of sound sources, while accounting for non-stationarity. I will also discuss the application of these models to various applications such as source separation, denoising, and content based audio processing, showing why they yield improved performance when compared to non-negative spectrogram factorization techniques.
-
- Date & Time: Thursday, September 6, 2012; 12:00 PM
Speaker: Dr. Daisuke Saito, The University of Tokyo
Research Area: Speech & Audio
Abstract - In voice conversion studies, realization of conversion from/to an arbitrary speaker's voice is one of the important objectives. For this purpose, eigenvoice conversion (EVC) based on an eigenvoice Gaussian mixture model (EV-GMM) was proposed. In the EVC, similarly to speaker recognition approaches, a speaker space is constructed based on GMM supervectors which are high-dimensional vectors derived by concatenating the mean vectors of each of the speaker GMMs. In the speaker space, each speaker is represented by a small number of weight parameters of eigen-supervectors. In this talk, we revisit construction of the speaker space by introducing the tensor analysis of training data set. In our approach, each speaker is represented as a matrix of which the row and the column respectively correspond to the Gaussian component and the dimension of the mean vector, and the speaker space is derived by the tensor analysis of the set of the matrices. Our approach can solve an inherent problem of supervector representation, and it improves the performance of voice conversion. Experimental results of one-to-many voice conversion demonstrate the effectiveness of the proposed approach.
-
- Date: March 31, 2012
Where: International Workshop on Statistical Machine Learning for Speech Processing (IWSML)
MERL Contact: Jonathan Le Roux
Research Area: Speech & Audio
Brief - The paper "Latent Dirichlet Reallocation for Term Swapping" by Heaukulani, C., Le Roux, J. and Hershey, J.R. was presented at the International Workshop on Statistical Machine Learning for Speech Processing (IWSML).
-
- Date: March 13, 2012
Where: Acoustical Society of Japan Spring Meeting (ASJ)
MERL Contact: Jonathan Le Roux
Research Area: Speech & Audio
Brief - The paper "Speech Enhancement by Indirect VTS" by Le Roux, J. and Hershey, J.R. was presented at the Acoustical Society of Japan Spring Meeting (ASJ).
-
- Date & Time: Wednesday, February 22, 2012; 11:00 AM
Speaker: Dr. Charles Cadieu, McGovern Institute for Brain Research, MIT
MERL Host: Jonathan Le Roux
Research Area: Speech & Audio
Abstract - The human visual system processes complex patterns of light into a rich visual representation where the objects and motions of our world are made explicit. This remarkable feat is performed through a hierarchically arranged series of cortical areas. Little is known about the details of the representations in the intermediate visual areas. Therefore, we ask the question: can we predict the detailed structure of the representations we might find in intermediate visual areas?
In pursuit of this question, I will present a model of intermediate-level visual representation that is based on learning invariances from movies of the natural environment and produces predictions about intermediate visual areas. The model is composed of two stages of processing: an early feature representation layer, and a second layer in which invariances are explicitly represented. Invariances are learned as the result of factoring apart the temporally stable and dynamic components embedded in the early feature representation. The structure contained in these components is made explicit in the activities of second-layer units that capture invariances in both form and motion. When trained on natural movies, the first-layer produces a factorization, or separation, of image content into a temporally persistent part representing local edge structure and a dynamic part representing local motion structure. The second-layer units are split into two populations according to the factorization in the first-layer. The form-selective units receive their input from the temporally persistent part (local edge structure) and after training result in a diverse set of higher-order shape features consisting of extended contours, multi-scale edges, textures, and texture boundaries. The motion-selective units receive their input from the dynamic part (local motion structure) and after training result in a representation of image translation over different spatial scales and directions, in addition to more complex deformations. These representations provide a rich description of dynamic natural images, provide testable hypotheses regarding intermediate-level representation in visual cortex, and may be useful representations for artificial visual systems.
-
- Date & Time: Thursday, October 20, 2011; 2:00 PM -5:00 PM
Location: MERL
MERL Contact: Jonathan Le Roux
Research Area: Speech & Audio
Brief - MERL is hosting a mini-symposium on audio and music signal processing, with three talks by eminent researchers in the field: Prof. Mark Plumbley, Dr. Cedric Fevotte and Prof. Nobutaka Ono.
-