ASR emotional speech: Clarifying the issues and enhancing performance

There are multiple reasons to expect that recognising the verbal content of emotional speech will be a difficult problem, and recognition rates reported in the literature are in fact low. Including information about prosody improves recognition rate for emotions simulated by actors, but its relevance to the freer patterns of spontaneous speech is unproven. This paper shows that recognition rate for spontaneous emotionally coloured speech can be improved by using a language model based on increased representation of emotional utterances. The models are derived by adapting an already existing corpus, the British National Corpus (BNC). An emotional lexicon is used to identify emotionally coloured words, and sentences containing these words are recombined with the BNC to form a corpus with a raised proportion of emotional material. Using a language model based on that technique improves recognition rate by about 20%. (c) 2005 Elsevier Ltd. All rights reserved.

The relationship between agricultural intensification and biological control: experimental tests across Europe

Agricultural intensification can affect biodiversity and related ecosystem services such as biological control, but large-scale experimental evidence is missing. We examined aphid pest populations in cereal fields under experimentally reduced densities of (1) ground-dwelling predators (-G), (2) vegetation-dwelling predators and parasitoids (-V), (3) a combination of (1) and (2) (-G-V),compared with open-fields (control), in contrasting landscapes with low vs. high levels of agricultural intensification (AI), and in five European regions. Aphid populations were 28%, 97%, and 199% higher in -G, -V, and -G -V treatments, respectively, compared to the open fields, indicating synergistic effects of both natural-enemy groups. Enhanced parasitoid : host and predator : prey ratios were related to reduced aphid population density and population growth. The relative importance of parasitoids and vegetation-dwelling predators greatly differed among European regions, and agricultural intensification affected biological control and aphid density only in some regions. This shows a changing role of species group identity in diverse enemy communities and a need to consider region-specific landscape management.

Increasing density of rare species of intertidal gastropods: tests of competitive ability compared with common species.

Many assemblages contain numerous rare species, which can show large increases in abundances. Common species can become rare. Recent calls for experimental tests of the causes and consequences of rarity prompted us to investigate competition between co-existing rare and common species of intertidal gastropods. In various combinations, we increased densities of rare gastropod species to match those of common species to evaluate effects of intra- and interspecific competition on growth and survival of naturally rare or naturally common species at small and large densities. Rarity per se did not cause responses of rare species to differ from those of common species. Rare species did not respond to the abundances of other rare species, nor show consistently different responses from those of common species. Instead, individual species responded differently to different densities, regardless of whether they are naturally rare or abundant. This type of experimental evidence is important to be able to predict the effects of increased environmental variability on rare as opposed to abundant species and therefore, ultimately, on the structure of diverse assemblages. © 2012 Inter-Research.


--------------------------------------------------------------------------------

Reaxys Database Information|

--------------------------------------------------------------------------------

Political Violence and Child Adjustment: Longitudinal Tests of Sectarian Antisocial Behavior, Family Conflict, and Insecurity as Explanatory Pathways

Understanding the impact of political violence on child maladjustment is a matter of international concern. Recent research has advanced a social ecological explanation for relations between political violence and child adjustment. However, conclusions are qualified by the lack of longitudinal tests. Toward examining pathways longitudinally, mothers and their adolescents (M = 12.33, SD = 1.78, at Time 1) from 2-parent families in Catholic and Protestant working class neighborhoods in Belfast, Northern Ireland, completed measures assessing multiple levels of a social ecological model. Utilizing autoregressive controls, a 3-wave longitudinal model test (T1, n = 299; T2, n = 248; T3, n = 197) supported a specific pathway linking sectarian community violence, family conflict, childrens insecurity about family relationships, and adjustment problems.

Creating a spontaneous conversational speech corpus

Speech recognition and language analysis of spontaneous speech arising in naturally spoken conversations are becoming the subject of much research. However, there is a shortage of spontaneous speech corpora that are freely available for academics. We therefore undertook the building of a natural conversation speech database, recording over 200 hours of conversations in English by over 600 local university students. With few exceptions, the students used their own cell phones from their own rooms or homes to speak to one another, and they were permitted to speak on any topic they chose. Although they knew that they were being recorded and that they would receive a small payment, their conversations in the corpus are probably very close to being natural and spontaneous. This paper describes a detailed case study of the problems we faced and the methods we used to make the recordings and control the collection of these social science data on a limited budget.

CLOSE—A Data-Driven Approach to Speech Separation

This paper studies single-channel speech separation, assuming unknown, arbitrary temporal dynamics for the speech signals to be separated. A data-driven approach is described, which matches each mixed speech segment against a composite training segment to separate the underlying clean speech segments. To advance the separation accuracy, the new approach seeks and separates the longest mixed speech segments with matching composite training segments. Lengthening the mixed speech segments to match reduces the uncertainty of the constituent training segments, and hence the error of separation. For convenience, we call the new approach Composition of Longest Segments, or CLOSE. The CLOSE method includes a data-driven approach to model long-range temporal dynamics of speech signals, and a statistical approach to identify the longest mixed speech segments with matching composite training segments. Experiments are conducted on the Wall Street Journal database, for separating mixtures of two simultaneous large-vocabulary speech utterances spoken by two different speakers. The results are evaluated using various objective and subjective measures, including the challenge of large-vocabulary continuous speech recognition. It is shown that the new separation approach leads to significant improvement in all these measures.