Michael L. Seltzer

Michael L. Seltzer

Auto-generated profile, based on publicly available information

Michael L. Seltzer

Microsoft AI and Research, Redmond, WA, USA

Research Focus & Metrics

Academic Career Length

25Years (2002-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
26 September 2017 in IEEE/ACM Transactions on Audio, Speech, and Language Processing
Verified
Journal Article
Closed access
WXWayne Xiong
Wayne Xiong
JDJasha Droppo
Jasha Droppo
XHXuedong Huang
Xuedong Huang
FSFrank Seide
Frank Seide
MSMichael L. Seltzer
Michael L. Seltzer
GZGeoffrey Zweig
Geoffrey Zweig
Conversational speech recognition has served as a flagship speech recognition task since the release of the Switchboard corpus in the 1990s. In this paper, we measure a human error rate on the widely used NIST 2000 test set for commercial bulk transcription. The error rate of professional transcribers is 5.9% for the Switchboard portion of the data, in which newly acquainted pairs of people discuss an assigned topic, and 11.3% for the CallHome portion, where friends and family members have open-ended conversations. In both cases, our automated system edges past the human benchmark, achieving error rates of 5.8% and 11.0%, respectively. The key to our system's performance is the use of various convolutional and long-short-term memory acoustic model architectures, combined with a novel spatial smoothing method and lattice-free discriminative acoustic training, multiple recurrent neural network language modeling approaches, and a systematic use of system combination. Comparing frequent errors in our human and machine transcripts, we find them to be remarkably similar, and highly correlated as a function of the speaker. Human subjects find it very difficult to tell which errorful transcriptions come from humans. Overall, this suggests that, given sufficient matched training data, conversational speech transcription engines are approximating human parity in both quantitative and qualitative terms.

1 May 2008
Verified
Conference Paper
Closed access
MSMichael L. Seltzer
Michael L. Seltzer
In this paper we describe two families of algorithms for hands-free speech recognition using microphone arrays. Enhancement-based approaches use a cascade of independent processing blocks to perform speech enhancement followed by speech recognition. We discuss the reasons why this approach may be sub-optimal and motivate the need for a solution that tightly integrates all processing blocks into a common unified framework. This leads to a second family of algorithms called unified approaches which considers all processing stages to be components of a single system that operates with the common goal of improved recognition accuracy. We describe several examples of such algorithms that have been shown to outperform more traditional signal-processing-based approaches. In doing so, we hope to convey the benefits of performing hands-free speech recognition in this manner and motivate further research in this area.

16 April 2007 in IEEE Signal Processing Letters
Verified
Journal Article
Closed access
ASAmarnag Subramanya
Amarnag Subramanya
MSMichael L. Seltzer
Michael L. Seltzer
AAAlex Acero
Alex Acero
Computers are increasingly being used to capture audio in various applications such as video conferencing and meeting recording. In many of these applications, user may be simultaneously typing on the keyboard, e.g., to take notes or search for information. As a result, the captured speech signals are significantly corrupted by sounds generated by the user's keystrokes. In this paper we propose an algorithm to automatically detect and remove keystrokes from speech signals. The proposed method does not require any user training or enrollment and is computationally efficient. The keystroke removal algorithm generates significantly enhanced speech as measured by both user listening tests and speech recognition experiments

1 September 2003
Verified
Conference Paper
Closed access
MSMichael L. Seltzer
Michael L. Seltzer
JDJasha Droppo
Jasha Droppo
AAAlex Acero
Alex Acero

1 May 2002
Verified
Conference Paper
Open access
MSMichael L. Seltzer
Michael L. Seltzer
BRBhiksha Raj
Bhiksha Raj
Richard M. SternRichard M. Stern
Richard M. Stern
We present a new array processing algorithm for microphone array speech recognition. Conventionally, the goal of array processing is to take distorted signals captured by the array and generate a cleaner output waveform. However, speech recognition systems operate on a set of features derived from the waveform, rather than the waveform itself. The goal of an array processor used in conjunction with a recognition system is to generate a waveform which produces a set of recognition features which maximize die likelihood for the words that are spoken, rather than to minimize the waveform distortion. We propose a new array processing algorithm which maximizes the likelihood of the recognition features. This is accomplished through the use of a new objective function which utilizes information from the recognition system itself, obtained in an unsupervised manner, to optimize the parameters of a filter-and-sum array processor. Using the proposed method, improvements in word error rate of up to 36% over conventional methods are achieved on real microphone array tasks in a wide range of environments.

8 March 2013 in arXiv
Verified
Preprint
Open access
PDF file
DYDong Yu
Dong Yu
MSMichael L. Seltzer
Michael L. Seltzer
JLJinyu Li
Jinyu Li
JHJui-Ting Huang
Jui-Ting Huang
FSFrank Seide
Frank Seide
Recent studies have shown that deep neural networks (DNNs) perform significantly better than shallow networks and Gaussian mixture models (GMMs) on large vocabulary speech recognition tasks. In this paper, we argue that the improved accuracy achieved by the DNNs is the result of their ability to extract discriminative internal representations that are robust to the many sources of variability in speech signals. We show that these representations become increasingly insensitive to small perturbations in the input with increasing network depth, which leads to better speech recognition performance with deeper networks. We also show that DNNs cannot extrapolate to test samples that are substantially different from the training examples. If the training data are sufficiently representative, however, internal features learned by the DNN are relatively stable with respect to speaker differences, bandwidth differences, and environment distortion. This enables DNN-based recognizers to perform as well or better than state-of-the-art systems based on GMMs or shallow networks without the need for explicit model adaptation or feature normalization.