Madhusudana Shashanka

Supervised and semi-supervised separation of sounds from single-channel mixtures

In this paper we describe a methodology for model-based single channel separation of sounds. We present a sparse latent variable model that can learn sounds based on their distribution of time/ frequency energy. This model can then be used to extract known types of sounds from mixtures in two scenarios. One being the case where all sound types in the mixture are known, and the other being being the case where only the target or the interference models are known. The model we propose has close ties to non-negative decompositions and latent variable models commonly used for semantic analysis.

Sparse and shift-invariant feature extraction from non-negative data

In this paper we describe a technique that allows the extraction of multiple local shift-invariant features from analysis of non-negative data of arbitrary dimensionality. Our approach employs a probabilistic latent variable model with sparsity constraints. We demonstrate its utility by performing feature extraction in a variety of domains ranging from audio to images and video.

Optimal skewed data allocation on multiple channels with flat broadcast per channel

Broadcast is an efficient and scalable way of transmitting data to an unlimited number of clients that are listening to a channel. Cyclically broadcasting data over the channel is a basic scheduling technique, which is known as flat scheduling. When multiple channels are available, a data allocation technique is needed to assign data to channels. Partitioning data among channels in an unbalanced way, depending on data popularities, is an allocation technique known as skewed allocation. The problem of data broadcasting over multiple channels is considered, assuming skewed data allocation to channels and flat data scheduling per channel, with the objective of minimizing the average waiting time of the clients. First, several algorithms, based on dynamic programming, are presented which provide optimal solutions for N data items and K channels. Specifically, for data items with uniform lengths, an O(NK log N) time algorithm is proposed, which improves over the previously known O(N/sup 2/K) time algorithm. When K/spl les/4, a simpler O(N log N) time algorithm is exhibited which requires only O(N) time if the data items are sorted. Moreover, for data items with nonuniform lengths, it is shown that the problem is NP-hard when K=2 and strong NP-hard for arbitrary K. In the former case, a pseudopolynomial algorithm is discussed whose time is O(NZ), where Z is the sum of the data lengths. In the latter case, an algorithm is devised with time exponential in the maximum data length, which can optimally solve, in reasonable time, only small instances. For larger instances, a new heuristic is devised which is experimentally tested on some benchmarks whose popularities are characterized by Zipf distributions. Such experimental tests reveal that the new heuristic proposed here always outperforms the best previously known heuristic in terms of solution quality.

Sparse Overcomplete Decomposition for Single Channel Speaker Separation

We present an algorithm for separating multiple speakers from a mixed single channel recording. The algorithm is based on a model proposed by Raj and Smaragdis (2005). The idea is to extract certain characteristic spectra-temporal basis functions from training data for individual speakers and decompose the mixed signals as linear combinations of these learned bases. In other words, their model extracts a compact code of basis functions that can explain the space spanned by spectral vectors of a speaker. In our model, we generate a sparse-distributed code where we have more basis functions than the dimensionality of the space. We propose a probabilistic framework to achieve sparsity. Experiments show that the resulting sparse code better captures the structure in data and hence leads to better separation.

Missing data imputation for spectral audio signals

With the recent attention to audio processing in the time -frequency domain we increasingly encounter the problem of missing data. In this paper we present an approach that allows for imputing missing values in the time-frequency domain of audio signals. The presented approach is able to deal with real-world polyphonic signals by performing imputation even in the presence of complex mixtures. We show that this approach outperforms generic imputation approaches, and we present a variety of situations that highlight its utility.

Model-based real-time whole building energy performance monitoring and diagnostics

Building energy systems often consume approximately 16% more energy [Mills, E. 2011. “Building Commissioning: A Golden Opportunity for Reducing Energy Costs and Greenhouse Gas Emissions in the United States.” Energy Efficiency 4 (2): 145–173] than is necessary due to system deviation from the design intent. Identifying the root causes of energy waste in buildings can be challenging largely because energy flows are generally invisible. To help address this challenge, we present a model-based, real-time whole building energy diagnostics and performance monitoring system. The proposed system continuously acquires performance measurements of heating, ventilation and air-conditioning, lighting and plug equipment usage and compare these measurements in real-time to a reference EnergyPlus model that either represents the design intent for the building or has been calibrated to represent acceptable performance. A proof-of-concept demonstration in a real building is also presented.

Missing Data Imputation for Time-Frequency Representations of Audio Signals

With the recent attention towards audio processing in the time-frequency domain we increasingly encounter the problem of missing data within that representation. In this paper we present an approach that allows us to recover missing values in the time-frequency domain of audio signals. The presented approach is able to deal with real-world polyphonic signals by operating seamlessly even in the presence of complex acoustic mixtures. We demonstrate that this approach outperforms generic missing data approaches, and we present a variety of situations that highlight its utility.

Latent Dirichlet Decomposition for Single Channel Speaker Separation

We present an algorithm for the separation of multiple speakers from mixed single-channel recordings by latent variable decomposition of the speech spectrogram. We model each magnitude spectral vector in the short-time Fourier transform of a speech signal as the outcome of a discrete random process that generates frequency bin indices. The distribution of the process is modeled as a mixture of multinomial distributions, such that the mixture weights of the component multinomials vary from analysis window to analysis window. The component multinomials are assumed to be speaker specific and are learned from training signals for each speaker. We model the prior distribution of the mixture weights for each speaker as a Dirichlet distribution. The distributions representing magnitude spectral vectors for the mixed signal are decomposed into mixtures of the multinomials for all component speakers. The frequency distribution, i.e the spectrum for each speaker, is reconstructed from this decomposition

Maximally Bijective Discretization for data-driven modeling of complex systems

Phase-space discretization is a necessary step for study of continuous dynamical systems using a language-theoretic approach. It is also critical for many machine learning techniques, e.g., probabilistic graphical models (Bayesian Networks, Markov models). This paper proposes a novel discretization method - Maximally Bijective Discretization, that finds a discretization on the dependent variables given a discretization on the independent variables such that the correspondence between input and output variables in the continuous domain is preserved in discrete domain for the given dynamical system.

Optimal multi-channel data allocation with flat broadcast per channel

Summary form only given. Broadcast is an efficient and scalable way of transmitting data to an unlimited number of clients that are listening to a channel. Cyclically broadcasting data over the channel is a basic scheduling technique, which is known as flat scheduling. When multiple channels are available, partitioning data among channels in an unbalanced way, depending on data popularities, is an allocation technique known as skewed allocation. In this paper, the problem of data broadcasting over multiple channels is considered assuming skewed data allocation to channels and fiat data scheduling per channel, with the objective of minimizing the average waiting time of the clients. Several algorithms, based on dynamic programming, are presented which provide optimal solutions for N data items and K channels. Specifically, for data items with uniform lengths, an O(NKlogN) time algorithm is proposed, which improves over the previously known O(N/sup 2/K) time algorithm. When K /spl les/ 4, faster O(N) time algorithms are exhibited. Moreover, for data items with nonuniform lengths, it is shown that the problem is NP-hard when K = 2, and strong NP-hard for arbitrary K. In the former case, a pseudo-polynomial algorithm is discussed, whose time is O(NZ) where Z is the sum of the data lengths.