ICASSP 2016

ICASSP is the world's largest and most comprehensive technical conference on signal processing and its applications. It provides a fantastic networking opportunity for like-minded professionals from around the world. ICASSP 2016 conference will feature world-class presentations by internationally renowned speakers and cutting-edge session topics.

Agreement and Disagreement Classification of Dyadic Interactions Using Vocal and Gestural Cues

Multimodal, Speech, Motion capture, Agreement Disagreement classification, JESTKOD, Human machine interaction, Dyadic interaction modeling, Negotiation

Agreement and Disagreement Classification of Dyadic Interactions Using Vocal and Gestural Cues-Presentation.pdf

Agreement and Disagreement Classification of Dyadic Interactions Using Vocal and Gestural Cues-Presentation.pdf (60)

Categories:: Multimedia human-machine interface and interaction

53 Views

Distributed Estimation of Latent Parameters in State Space Models Using Separable Likelihoods

Separable likelihoods in sensor self-localisation using opportunistic measurements

This work is a part of our research on scalable and/or distributed fusion and sensor calibration. We address parameter estimation in multi-sensor state space models which underpins surveillance applications with sensor networks. The parameter likelihood of the problem involves centralised Bayesian filtering of multi-sensor data, which lacks scalability with the number of sensors and induces a large communication load. We propose separable likelihoods which approximate the centralised likelihood with single sensor filtering terms.

icassp2016-muney-Presentation_0.1.pdf

icassp2016-muney-Presentation_0.1.pdf (665)

Categories:: Signal and System Modeling, Representation and Estimation
Statistical Signal Processing

4 Views

Discriminatively Trained Joint Speaker and Environment Representations for Adaptation of Deep Neural Network Acoustic Models

mfy-icassp16-slides.pdf

mfy-icassp16-slides.pdf (603)

Categories:: Acoustic Modeling for Automatic Speech Recognition (SPE-RECO)

16 Views

KEYWORD SEARCH USING QUERY EXPANSION FOR GRAPH-BASED RESCORING OF HYPOTHESIZED DETECTIONS

In this work, we propose a novel framework for rescoring keyword search (KWS) detections using acoustic samples extracted from the training data. We view the keyword rescoring task as an information retrieval task and adopt the idea of query expansion. We expand a textual keyword with multiple speech keyword samples extracted from the training data. In this way, the hypothesized detections are compared with the multiple keywords using non-parametric approaches such as dynamic time warping (DTW).