Improvement of rejection performance of keyword spotting using anti-keywords derived from large vocabulary considering acoustical similarity to keywords

A Russian Keyword Spotting System Based on Large Vocabulary Continuous Speech Recognition and Linguistic Knowledge

Journal of Electrical and Computer Engineering ◽

10.1155/2016/4062786 ◽

2016 ◽

Vol 2016 ◽

pp. 1-9 ◽

Cited By ~ 2

Author(s):

Valentin Smirnov ◽

Dmitry Ignatov ◽

Michael Gusev ◽

Mais Farkhadov ◽

Natalia Rumyantseva ◽

...

Keyword(s):

Speech Recognition ◽

Real Life ◽

Training Data ◽

Linguistic Knowledge ◽

Keyword Spotting ◽

Continuous Speech ◽

Continuous Speech Recognition ◽

Large Vocabulary ◽

Effective Combination ◽

Industrial Use

The paper describes the key concepts of a word spotting system for Russian based on large vocabulary continuous speech recognition. Key algorithms and system settings are described, including the pronunciation variation algorithm, and the experimental results on the real-life telecom data are provided. The description of system architecture and the user interface is provided. The system is based on CMU Sphinx open-source speech recognition platform and on the linguistic models and algorithms developed by Speech Drive LLC. The effective combination of baseline statistic methods, real-world training data, and the intensive use of linguistic knowledge led to a quality result applicable to industrial use.

Download Full-text

A fast algorithm for large vocabulary keyword spotting application

IEEE Transactions on Speech and Audio Processing ◽

10.1109/89.294361 ◽

1994 ◽

Vol 2 (3) ◽

pp. 449-452 ◽

Cited By ~ 9

Author(s):

Eng-Fong Huang ◽

Hsiao-Chuan Wang ◽

F.K. Soong

Keyword(s):

Fast Algorithm ◽

Keyword Spotting ◽

Large Vocabulary

Download Full-text

A word-length-dependent confidence measure for large vocabulary Chinese keyword spotting

TENCON '97 Brisbane - Australia. Proceedings of IEEE TENCON '97. IEEE Region 10 Annual Conference. Speech and Image Technologies for Computing and Telecommunications (Cat. No.97CH36162) ◽

10.1109/tencon.1997.648276 ◽

2002 ◽

Author(s):

Bo-Ren Bai ◽

Hsin-Min Wang ◽

Lin-Shan Lee

Keyword(s):

Word Length ◽

Keyword Spotting ◽

Confidence Measure ◽

Large Vocabulary

Download Full-text

A New Framework For Large Vocabulary Keyword Spotting Using Two-Pass Confidence Measure

The Proceedings of the Multiconference on "Computational Engineering in Systems Applications" ◽

10.1109/cesa.2006.4281625 ◽

2006 ◽

Cited By ~ 4

Author(s):

Yingna Chen ◽

Tao Hou ◽

Sha Meng ◽

Shan Zhong ◽

Jia Liu

Keyword(s):

Keyword Spotting ◽

Confidence Measure ◽

Large Vocabulary ◽

New Framework

Download Full-text

Keyword-spotting using SRI's DECIPHER large-vocabulary speech-recognition system

IEEE International Conference on Acoustics Speech and Signal Processing ◽

10.1109/icassp.1993.319341 ◽

1993 ◽

Cited By ~ 27

Author(s):

M. Weintraub

Keyword(s):

Speech Recognition ◽

Recognition System ◽

Keyword Spotting ◽

Speech Recognition System ◽

Large Vocabulary ◽

Large Vocabulary Speech Recognition

Download Full-text

Aversive Textures and Their Role in Food Rejection

10.31234/osf.io/quk2f ◽

2020 ◽

Author(s):

Robert Pellegrino ◽

Curtis Luckett

Keyword(s):

Cluster Analysis ◽

Psychological Factors ◽

Sensory Attributes ◽

Prominent Feature ◽

Large Vocabulary ◽

Common Food ◽

Sensory Attribute ◽

Touch Sensitivity

Texture is a prominent feature in foods and consequently can be the reason a food is accepted or rejected. However, other sensory attributes, such as flavor/taste, aroma, sound and appearance may also lead to the rejection of food and motivations other than unpleasantness exist in unacceptance. To date, these motivations for food rejection have been studied in isolation and their relationships with psychological factors have not been tested. This study measured reasons people reject a food and probed into the specifics of texture rejection. A large U.S. sample (N=473) was asked to rate their motivations for rejecting a food, list foods that were disliked due to unpleasant sensory attributes, specify the unpleasant sensory attribute(s), and complete an assessment of general touch sensitivity. Results showed 94% of individuals reject a food due to its texture, a rate comparable to flavor-based rejection. Looking at the number of foods being rejected, flavor was the most common food attribute, followed by texture and then aroma. From a linguistic standpoint, aversive textures encompass a large vocabulary, larger than liked textures, and the same food may be rejected due to a single or combination of texture terms. Viscosity (e.g. slimy) and hardness (e.g. mushy) are the most common aversive texture types, but through cluster analysis subsets of individuals were identified that are more aversive to other textures. This study emphasizes the role of aversive textures in food rejection and provides many avenues for future investigations.

Download Full-text