Ove Kjeld Andersen

EcoSky: Reducing vehicular environmental impact through eco-routing

Reduction in greenhouse gas emissions from transportation attracts increasing interest from governments, fleet managers, and individual drivers. Eco-routing, which enables drivers to use eco-friendly routes, is a simple and effective approach to reducing emissions from transportation. We present EcoSky, a system that annotates edges of a road network with time dependent and uncertain eco-weights using GPS data and that supports different types of eco-routing. Basic eco-routing returns the most eco-friendly routes; skyline eco-routing takes into account not only fuel consumption but also travel time and distance when computing eco-routes; and personalized eco-routing considers each drivers past behavior and accordingly suggests different routes to different drivers.

On the robust automatic segmentation of spontaneous speech

The results from applying an improved algorithm to the task of automatic segmentation of spontaneous telephone quality speech are presented, and compared to the results from those resulting from superimposing white noise. Three segmentation algorithms are compared which are all based on variants of the Spectral Variation Function. Experimental results are obtained on the OGI multi language telephone speech corpus (OGLTS). We show that the use of the auditory forward and backward masking effects prior to the SVF computation increases the robustness of the algorithm to white noise. When the average signal to noise ratio (SNR) is decreased to 10 dB, the peak ratio (defined as the ratio of the number of peaks measured at the target over the original SNRs) is increased by 16%, 12%, and 11% for the MFC (Mel Frequency Cepstra), RASTA (Relative Spectral Processing), and the FBDYN (Forward Backward Auditory Masking Dynamic Cepstra) SVF segmentation algorithms, respectively.

Employing heterogeneous information in a multi-stream framework

A multi-stream speech recogniser is based on the combination of multiple feature streams each containing complementary information. In the past, multi-stream research has typically focused on systems that use a single feature extraction method. This heritage from conventional speech recognisers is an unnecessary restriction and both psychoacoustic and phonetic knowledge strongly motivate the use of heterogeneous features. In this paper we investigate how heterogeneous processing can be used in two different multi-stream configurations: first, a system where each stream handles a different frequency region of the speech (a multi-band recogniser) and, second a multi-stream recogniser where each stream handles the full frequency region. For each type of system we compare the performance using both homogeneous and heterogeneous processing. We demonstrate that the use of heterogeneous information significantly improves the clean speech recognition performance motivating us to continue exploring more specifically designed stream processing.

Multi-lingual label alignment using acoustic-phonetic features derived by neural-network technique

In previous work on label alignment, encouraging results were obtained using selected acoustic-phonetic features to model the individuals speech phonemes. Selection was based on minimal covariance between features on the one hand, and the inclusion of features underlying critical phonological opposition on the other. In the present work, principal component analysis was applied to give a number of uncorrelated output parameters which maximally exploit the discriminatory power of the features and are derived independently of the phonological functionality. Results of label alignment on three different European languages, Danish, English, and Italian, using different numbers of principal parameters show that the accuracy with ten parameters is at least as good as with 15 manually selected features. The best result is found for British English, which has 78% of its phoneme transition boundaries positioned within +or-20 ms of manually placed reference boundaries.<<ETX>>

Language-identification using language-dependent phonemes and language-independent speech units

The paper reports on results from ongoing research on language identification (LID) performed on the three languages: American-English, German and Spanish. The speech material used is from the Oregon Graduate Institute Spontaneous Telephone Speech Corpus, OGI-TS. The baseline LID system consists of three parallel phoneme recognisers, each of which are followed by three language modelling modules each characterising the bigram probabilities. The phoneme models used are derived on the basis of the combined speech corpus comprising the three languages. The phonemes are handled differently in analysis performed in two experiments. In the first experiment they are trained and tested language specifically. In the second, they are separated into a number of groups, one of which contains those language independent speech units which are similar enough to be equated across the training languages, the remaining containing the non combinable language dependent phonemes for each of the languages. A data driven technique has been devised to separate the speech sounds contained within the training corpus into these groups. In order to prepare for an optimal separation between the input classes, a linear discriminant analysis is performed on the training speech material. Results from a number of experiments show that average language identification scores of close to 90% can be retained by the LID system presented here, even for a high number of language independent speech units.

Separation of mixed phase signals by zeros of the z-transform - A reformulation of complex cepstrum based separation by causality

In recent studies, a non-parametric speech waveform representation (rep.) based on zeros of the z-transform (ZZT) has been proposed. The ZZT rep. has successfully been applied in separating mixed phase signals, e.g. pitch-synchronously windowed speech, into min/max phase by using the unit circle as discriminant. As the ZZT rep. is obtained by factorization of the z-transform, relations to the complex cepstrum (CC) exist. The present paper interrelates the ZZT rep. with the CC via factorization of the z-transform, and demonstrates that unit circle discrimination of a ZZT rep. can be formulated as a CC based separation by causality. A numerical experiment supplements theory by separating a range of LF glottal flow waveforms into their opening and closing phase constituents. Further, randomized mixed phase sequences are separated. As the CC based separation also can be obtained via FFT it has a lower time and space complexity than the ZZT based counterpart.

On the use of acoustic-phonetic features in interactive labelling of multi-lingual speech corpora

Results are reported from research on the use of continuously valued acoustic-phonetic features in the multi-language label alignment of combined speech corpora from three European languages: Danish, English, and Italian. A self-organizing neural network is used to transform cepstrum coefficients into a set of features, which are subsequently transformed into a set of principal components. These are used to model individual phonemes, which are used in a Viterbi search/level-building process to align an independently given string of phonemes with the corresponding speech signal. The results obtained show an overall accuracy of 55.7% in the positioning of the label boundary transitions in the combined test corpus. Detailed analysis shows that certain sound class boundaries are very accurately positioned. To provide a general solution to the problem of positioning badly positioned boundaries, an interactive component of the alignment system has been developed. First results demonstrate this component to be very valuable in the task of user-assisted boundary positioning.<<ETX>>

CO 2 NNIE: personalized fuel consumption and CO 2 emissions

We propose a system for calculating the personalized annual fuel consumption and CO2 emissions from transportation. The system, named CO2NNIE, estimates the fuel consumption on the fastest route between the frequent destinations of the user. The travel time and fuel consumption estimated are based on 3.8 billion GPS records from 16 thousand cars and 198 million records from 218 cars annotated with fuel consumption data, respectively. The fuel consumption estimates from the system are validated using fuel-pump data. We find that estimates have good accuracy, i.e., are generally within 10% of the actual fuel consumption (4.6% deviation on average). We conclude, that the system provides new detailed information on CO2 emissions and fuel consumption for any make and model.

Introducing Phonetically Motivated, Heterogeneous Information into Automatic Speech Recognition

This chapter investigates a way to introduce more heterogeneous information into an existing ASR system. A phonetic expert is implemented which is specifically targeted at correcting the errors made by an existing ASR system. This gives a heterogeneous system, where the individual items are designed to be complementary. To avoid the curse of dimensionality problem, the expert information is introduced at the level of the acoustic model. Two types of expert configurations are used, each providing discriminative information regarding groups of phonetically related phonemes. The phonetic expert is implemented using an MLP. Experiments show that, when using the expert in conjunction with both a fullband and a multiband system speech recognition performance is increased and noise robustness improved for a range of noise levels.

Comparative performance analysis of statistical trajectory models in cellular environment

Two systems (statistical trajectory models (STM) and continuous density HMMs) utilizing three preprocessing methodologies (MFCC, RASTA and FBDYN) were evaluated on two databases, namely CTIMIT and the corresponding down-sampled TIMIT. Within the bounds of the experimental setup the comparative performance analysis showed that the STM significantly outperforms the HMM system on the CTIMIT database. Specifically, the performance of the STM system was found to be at least 10% better as compared to the one obtained by HMM when the RASTA preprocessing was used. The performance of both systems with FBDYN parametrization was found to be inferior to those using MFCC and RASTA. On the other hand, in low-noise conditions on the TIMIT database FBDYN yielded an improved performance for the HMM system, whereas STM achieved the best results with the MFCC parametrization.