Conventional automatic speech recognition (ASR) often neglects the spectral phase information in its front-end and feature extraction stages. The aim of this paper is to show the impact that enhancement of the noisy spectral phase has on ASR accuracy when dealing with speech signals corrupted with additive noise. Apart from proof-of-concept experiments using clean spectral phase, we also present a phase enhancement method as a phase-aware front-end and modified group delay as a phaseaware feature extractor, and the combination thereof. In experiments, we demonstrate the improved performance for each individual component and their combination, compared to the conventional phase-unaware Mel Frequency Cepstral Coefficients (MFCCs)-based ASR. We observe that the estimated phase information used in the front-end or feature extraction component improves the ASR word accuracy rate (WAR) by 20.98 % absolute for noise corrupted speech (averaged over SNRs ranging from 0 to 20 dB).
Uncertainty is ubiquitous in natural human communication. Human listeners assess the speaker's degree of uncertainty at any time in communication and use this information to shape dialogue. In contrast, currently available computer systems dealing with spoken language are usually not built to perform this task. The ability to detect uncertainty would likely lead to more natural human-computer dialogue. In order to detect uncertainty automatically, we extract linguistic, paralinguistic and dialogue-related features from the Kiel Corpus, a corpus of naturalistic task-oriented spoken German. We then use these features to train a random forests model. Our experimental results show that relatively high classification accuracy can be obtained while employing only 64 well-chosen features (73% accuracy, 69% F1). To our best knowledge, this is the first study of automatic uncertainty detection using German speech data as well as the first achieving good performance on everyday speech.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.