Frequency-based interpolation of sampled signals with applications in audio restoration

A new interpolation technique has been developed for replacing missing samples in a sampled waveform drawn from a stationary stochastic process, given the power spectrum for the process. The method works with a finite block of data and is based on the assumption that components of the block DFT are Gaussian zero-mean independent random variables with variance proportional to the power spectrum at each frequency value. These assumptions make the interpolator particularly suitable for signals with a sharply-defined harmonic structure, such as audio waveforms recorded from music or voiced speech. Some results are presented and comparisons are made with existing techniques.

Spoken document representations for probabilistic retrieval

This paper presents some developments in query expansion and document representation of our spoken document retrieval system and shows how various retrieval techniques affect performance for different sets of transcriptions derived from a common speech source. Modifications of the document representation are used, which combine several techniques for query expansion, knowledge-based on one hand and statistics-based on the other. Taken together, these techniques can improve Average Precision by over 19% relative to a system similar to that which we presented at TREC-7. These new experiments have also confirmed that the degradation of Average Precision due to a word error rate (WER) of 25% is quite small (3.7% relative) and can be reduced to almost zero (0.2% relative). The overall improvement of the retrieval system can also be observed for seven different sets of transcriptions from different recognition engines with a WER ranging from 24.8% to 61.5%. We hope to repeat these experiments when larger document collections become available, in order to evaluate the scalability of these techniques.

Model-based approaches to handling additive noise in reverberant environments

Model-based approaches to handle additive and convolutional noise have been extensively investigated and used. However, the application of these schemes to handling reverberant noise has received less attention. This paper examines the extension of two standard additive/convolutional noise approaches to handling reverberant noise. The first is an extension of vector Taylor series (VTS) compensation, reverberant VTS, where a mismatch function including reverberant noise is used. The second scheme modifies constrained MLLR to allow a wide-span of frames to be taken into account and projected into the required dimensionality. To allow additive noise to be handled, both these schemes are combined with standard VTS. The approaches are evaluated and compared on two tasks, MC-WSJ-AV, and a reverberant simulated version of AURORA-4. © 2011 IEEE.