In this work, we extend the TBC method, proposing a new similarity metric for selecting training data that results in significant gains over the one proposed in the original work, a new option that…
Speech & natural language publications
Resilient Data Augmentation Approaches to Multimodal Verification in the News Domain
Building on multimodal embedding techniques, we show that data augmentation via two distinct approaches improves results: entity linking and cross-domain local similarity scaling.
Natural Language Access: When Reasoning Makes Sense
We argue that to use natural language effectively, we must have both a deep understanding of the subject domain and a general-purpose reasoning capability.
Wideband Spectral Monitoring Using Deep Learning
We present a system to perform spectral monitoring of a wide band of 666.5 MHz, located within a range of 6 GHz of Radio Frequency (RF) bandwidth, using state-of-the-art deep learning approaches.
Dual orexin and MCH neuron-ablated mice display severe sleep attacks and cataplexy
Orexin/hypocretin-producing and melanin-concentrating hormone-producing (MCH) neurons are co-extensive in the hypothalamus and project throughout the brain to regulate sleep/wakefulness.
Mapping Individual to Group Level Collaboration Indicators Using Speech Data
Automatic detection of collaboration quality from the students’ speech could support teachers in monitoring group dynamics, diagnosing issues, and developing pedagogical intervention plans.
Analysis of Complementary Information Sources in the Speaker Embeddings Framework
Deep neural network (DNN)-based speaker embeddings have resulted in new, state-of-the-art text-independent speaker recognition technology. However, very limited effort has been made to understand DNN speaker embeddings. In this study, our aim is analyzing the behavior of the speaker recognition systems based on speaker embeddings toward different front-end features, including the standard Mel frequency cepstral coefficients (MFCC), as well as power normalized cepstral coefficients (PNCC), and perceptual linear prediction (PLP). Using a speaker recognition system based on DNN speaker embeddings and probabilistic linear discriminant analysis (PLDA), we compared different approaches to leveraging complementary information using score-, embeddings-, and feature-level combination. We report our results for Speakers in the Wild (SITW) and NIST SRE 2016 datasets. We found that first and second embeddings layers are complementary in nature. By applying score and embedding-level fusion we demonstrate relative improvements in equal error rate of 17% on NIST SRE 2016 and 10% on SITW over the baseline system.
Robust Speaker Recognition from Distant Speech under Real Reverberant Environments Using Speaker Embeddings
This article focuses on speaker recognition using speech acquired using a single distant or far-field microphone in an indoors environment. This study differs from the majority of speaker recognition research, which focuses on speech acquisition over short distances, such as when using a telephone handset or mobile device or far-field microphone arrays, for which beamforming can enhance distant speech signals. We use two large-scale corpora collected by retransmitting speech data in reverberant environments with multiple microphones placed at different distances. We first characterize three different speaker recognition systems ranging from a traditional universal background model (UBM) i-vector system to a state-of-the-art deep neural network (DNN) speaker embedding system with a probabilistic linear discriminant analysis (PLDA) back-end. We then assess the impact of microphone distance and placement, background noise, and loudspeaker orientation on the performance of speaker recognition system for distant speech data. We observe that the recently introduced DNN speaker embedding based systems are far more robust compared to i-vector based systems, providing a significant relative improvement of up to 54% over the baseline UBM i-vector system, and 45.5% over prior DNN-based speaker recognition technology.
Structure-based lead optimization to improve antiviral potency and ADMET properties of phenyl-1H-pyrrole-carboxamide entry inhibitors targeted to HIV-1 gp120
We are continuing our concerted effort to optimize our first lead entry antagonist, NBD-11021, which targets the Phe43 cavity of the HIV-1 envelope glycoprotein gp120, to improve antiviral potency and ADMET properties. In this report, we present a structure-based approach that helped us to generate working hypotheses to modify further a recently reported advanced lead entry antagonist, NBD-14107, which showed significant improvement in antiviral potency when tested in a single-cycle assay against a large panel of Env-pseudotyped viruses. We report here the synthesis of twenty-nine new compounds and evaluation of their antiviral activity in a single-cycle and multi-cycle assay to derive a comprehensive structure-activity relationship (SAR). We have selected three inhibitors with the high selectivity index for testing against a large panel of 55 Env-pseudotyped viruses representing a diverse set of clinical isolates of different subtypes. The antiviral activity of one of these potent inhibitors, 55 (NBD-14189), against some clinical isolates was as low as 63 nM. We determined the sensitivity of CD4-binding site mutated-pseudoviruses to these inhibitors to confirm that they target HIV-1 gp120. Furthermore, we assessed their ADMET properties and compared them to the clinical candidate attachment inhibitor, BMS-626529. The ADMET data indicate that some of these new inhibitors have comparable ADMET properties to BMS-626529 and can be optimized further to potential clinical candidates.