Speaker verification based on the fusion of speech acoustics and inverted articulatory signals

We propose apractical, feature-level and score-level fusion approach by combining acoustic and estimated articulatory information for both text independent and text dependent speaker verification. From a practical point of view, we study how to improve speaker verification performance by combining dynamic articulatory information with the conventional acoustic features. On text independent speaker verification, we find that concatenating articulatory features obtained from measured speech production data with conventional Mel-frequency cepstral coefficients (MFCCs) improves the performance dramatically. However, since directly measuring articulatory data is not feasible in many real world applications, we also experiment with estimated articulatory features obtained through acoustic-to-articulatory inversion. We explore both feature level and score level fusion methods and find that the overall system performance is significantly enhanced even with estimated articulatory features. Such a performance boost could be due to the inter-speaker variation information embedded in the estimated articulatory features. Since the dynamics of articulation contain important information, we included inverted articulatory trajectories in text dependent speaker verification. We demonstrate that the articulatory constraints introduced by inverted articulatory features help to reject wrong password trials and improve the performance after score level fusion. We evaluate the proposed methods on the X-ray Microbeam database and the RSR 2015 database, respectively, for the aforementioned two tasks. Experimental results show that we achieve more than 15% relative equal error rate reduction for both speaker verification tasks. (C) 2015 Elsevier Ltd. All rights reserved.

Botulinum neurotoxin type-A enters a non-recycling pool of synaptic vesicles

Neuronal communication relies on synaptic vesicles undergoing regulated exocytosis and recycling for multiple rounds of fusion. Whether all synaptic vesicles have identical protein content has been challenged, suggesting that their recycling ability may differ greatly. Botulinum neurotoxin type-A (BoNT/A) is a highly potent neurotoxin that is internalized in synaptic vesicles at motor nerve terminals and induces flaccid paralysis. Recently, BoNT/A was also shown to undergo retrograde transport, suggesting it might enter a specific pool of synaptic vesicles with a retrograde trafficking fate. Using high-resolution microscopy techniques including electron microscopy and single molecule imaging, we found that the BoNT/A binding domain is internalized within a subset of vesicles that only partially co-localize with cholera toxin B-subunit and have markedly reduced VAMP2 immunoreactivity. Synaptic vesicles loaded with pHrodo-BoNT/A-Hc exhibited a significantly reduced ability to fuse with the plasma membrane in mouse hippocampal nerve terminals when compared with pHrodo-dextran-containing synaptic vesicles and pHrodo-labeled anti-GFP nanobodies bound to VAMP2-pHluorin or vGlut-pHluorin. Similar results were also obtained at the amphibian neuromuscular junction. These results reveal that BoNT/A is internalized in a subpopulation of synaptic vesicles that are not destined to recycle, highlighting the existence of significant molecular and functional heterogeneity between synaptic vesicles.

Differential dynamics of the serotonin(1A) receptor in membrane bilayers of varying cholesterol content revealed by all atom molecular dynamics simulation

The serotonin(1A) receptor belongs to the superfamily of G protein-coupled receptors (GPCRs) and is a potential drug target in neuropsychiatric disorders. The receptor has been shown to require membrane cholesterol for its organization, dynamics and function. Although recent work suggests a close interaction of cholesterol with the receptor, the structural integrity of the serotonin(1A) receptor in the presence of cholesterol has not been explored. In this work, we have carried out all atom molecular dynamics simulations, totaling to 3s, to analyze the effect of cholesterol on the structure and dynamics of the serotonin(1A) receptor. Our results show that the presence of physiologically relevant concentration of membrane cholesterol alters conformational dynamics of the serotonin(1A) receptor and, on an average lowers conformational fluctuations. Our results show that, in general, transmembrane helix VII is most affected by the absence of membrane cholesterol. These results are in overall agreement with experimental data showing enhancement of GPCR stability in the presence of membrane cholesterol. Our results constitute a molecular level understanding of GPCR-cholesterol interaction, and represent an important step in our overall understanding of GPCR function in health and disease.

Stalking influenza by vaccination with pre-fusion headless HA mini-stem

Inaccuracies in prediction of circulating viral strain genotypes and the possibility of novel reassortants causing a pandemic outbreak necessitate the development of an anti-influenza vaccine with increased breadth of protection and potential for rapid production and deployment. The hemagglutinin (HA) stem is a promising target for universal influenza vaccine as stem-specific antibodies have the potential to be broadly cross-reactive towards different HA subtypes. Here, we report the design of a bacterially expressed polypeptide that mimics a H5 HA stem by protein minimization to focus the antibody response towards the HA stem. The HA mini-stem folds as a trimer mimicking the HA prefusion conformation. It is resistant to thermal/chemical stress, and it binds to conformation-specific, HA stem-directed broadly neutralizing antibodies with high affinity. Mice vaccinated with the group 1 HA mini-stems are protected from morbidity and mortality against lethal challenge by both group 1 (H5 and H1) and group 2 (H3) influenza viruses, the first report of cross-group protection. Passive transfer of immune serum demonstrates the protection is mediated by stem-specific antibodies. Furthermore, antibodies indudced by these HA stems have broad HA reactivity, yet they do not have antibody-dependent enhancement activity.