Low resource point process models for keyword spotting using unsupervised online learning

Abstract

Abstract-Point Process Models (PPM) have been widely used for keyword spotting applications. Training these models typically requires a considerable number of keyword examples. In this work, we consider a scenario where very few keyword examples are available for training. The availability of a limited number of training examples results in a PPM with poorly learnt parameters. We propose an unsupervised online learning algorithm that starts from a poor PPM model and updates the PPM parameters using newly detected samples of the keyword in a corpus under consideration and uses the updated model for further keyword detection. We test our algorithm on eight keywords taken from the TIMIT database, the training set of which, on average, has 469 samples of each keyword. With an initial set of only five samples of a keyword (corresponds to ∼ 1% of the total number of samples) followed by the proposed online parameter updating throughout the entire TIMIT train set, the performance on the TIMIT test set using the final model is found to be comparable to that of a PPM trained with all the samples of the respective keyword available from the entire TIMIT train set.

Voice Activation Systems for Embedded Devices: Systematic Literature Review

“…For more information on the proposed see Jansen and Niyogi (2009c), Jansen and Niyogi (2009b). Sadhu and Ghosh (2017) describe how to apply this approach in systems with limited resources using unsupervised online learning.…”

“…FOM (Gish et al, 1990;Rose and Paul, 1990;Naylor et al, 1992;Zeppenfeld and Waibel, 1992;Chang and Lippmann, 1994;Gish and Ng, 1993;Rohlicek et al, 1993;Knill and Young, 1996;Junkawitsch et al, 1997;Zheng et al, 1999;Szöke et al, 2005;Lehtonen, 2005;Jansen and Niyogi, 2009a,c;Szöke et al, 2010;Tabibian et al, 2011;Bohac, 2012;Sangeetha and Jothilakshmi, 2014;Sadhu and Ghosh, 2017;Tabibian et al, 2018) EER (Szöke et al, 2010;Bohac, 2012) Accuracy (Morgan et al, 1990Ida and Yamasaki, 1998;Ge and Yan, 2017;Benisty et al, 2018;Fernández-Marqués et al, 2018) FA/kw/h (Rohlicek et al, 1989;Vroomen and Normandin, 1992;Feng and Mazor, 1992;Leow et al, 2012;Kavya and Karjigi, 2014) ROC (Marcus, 1992;Siu et al, 1994;Keshet et al, 2009;Wöllmer et al, 2009bWöllmer et al, , 2013Shokri et al, 2013;Sadhu and Ghosh, 2017;Kumatani et al, 2017) Detection rate (Feng and Mazor, 1992;Khne et al, 2004;…”

Improving the effectiveness of voice activation systems with machine learning methods

“…For more information on the proposed, see Jansen and Niyogi (2009b). Sadhu and Ghosh (2017) described how to apply this approach in systems with limited resources using unsupervised online learning.…”