Looking for words that might cue backchannel feedback, I did a statistical analysis of the interlocutors’ words preceding 3363 instances of uh-huh in the Switchboard corpus. No clear cueing words were found, but collateral findings include the existence of semantic classes that slightly increase the likelihood of an upcoming uh-huh, the fact that different classes have their effects at differen...

In this paper we describe a new testbed for developing speech recognition algorithms a VoiceMail transcription task, analogous to other tasks such as the Switchboard, CallHome, and the Hub 4 tasks, which are currently used by speech recognition researchers. We describe the collection and use of a new VoiceMail database (that is available to the research community through the LDC), and also desc...

In this paper, the likelihood ratio approach is applied for measuring evidence provided by record data in favor of a hypothesis against an alternative under a random sampling scheme. Explicit expressions for probabilities of observing strong andweak evidences are derived. Asymptotic behaviors of these probabilities are investigated in a greater detail. Optimal sample size in which the substanti...

We describe a new method for pruning in dynamic models based on running an adaptive filtering algorithm online during decoding to predict aspects of the scores in the near future. These predictions are used to make well-informed pruning decisions during model expansion. We apply this idea to the case of dynamic graphical models and test it on a speech recognition database derived from Switchboa...

Syntactic persistence (Bock, 1986; Pickering and Branigan, 1998) is the tendency for speakers to repeat a syntactic structure that they have processed previously. In the following example (1) from the Switchboard corpus (Godfrey et al., 1992), the speaker produces the double object ditransitive structure, or NPNP (1-a), as opposed to the prepositional object structure, or NPPP (1-b), after havi...

Neuronal migration is a fundamental process of mammalian brain development. In migrating neurons, the nuclear membrane protein Nesprin-2 has been shown to serve as an adaptor pull nucleus along microtubule tracks. Current evidence that binds both minus-end-directed motor dynein well plus-end-directed kinesin. However, translocation neuronal long thought be primarily driven by motors. Intriguing...

We show that a deep convolutional network with an architecture inspired by the models used in image recognition can yield accuracy similar to a long-short term memory (LSTM) network, which achieves the state-of-the-art performance on the standard Switchboard automatic speech recognition task. Moreover, we demonstrate that merging the knowledge in the CNN and LSTM models via model compression fu...

This paper presents an experimental study of a maximum likelihood (ML) approach to irrelevant variability normalization (IVN) based training and unsupervised online adaptation for large vocabulary continuous speech recognition. A movingwindow based frame labeling method is used for acoustic sniffing. The IVN-based approach achieves a 10% relative word error rate reduction over an ML-trained bas...

This paper presents an unsupervised method for clustering spontaneous speech documents. The approach uses a hierarchical algorithm to automatically determine the number of clusters and a starting model for a subsequent iterative algorithm. We have evaluated this method on the Switchboard corpus and compared it to a set of supervised and other unsupervised methods. The results show that our meth...