Patent
System and method for measuring confusion among words in an adaptive speech recognition system
TLDR
In this article, a model-based approach is proposed for measuring confusability or similarity between given entry pairs, including text string pairs and acoustic model pairs, in systems such as speech recognition and synthesis systems.Abstract:
A system and method are proposed for measuring confusability or similarity between given entry pairs, including text string pairs and acoustic model pairs, in systems such as speech recognition and synthesis systems. A string edit distance (Levenshiten distance) can be applied to measure distance between any pair of text strings. It also can be used to calculate a confusion measurement between acoustic model pairs of different words and a model-driven method can be used to calculate a HMM model confusion matrix. This model-based approach can be efficiently calculated with low memory and low computational resources. Thus it can improve the speech recognition performance and models trained from text corpus.read more
Citations
More filters
Patent
System and method for open speech recognition
TL;DR: In this article, the authors present systems, methods and non-transitory computer-readable media for performing speech recognition across different applications or environments without model customization or prior knowledge of the received speech.
Patent
Realtime acoustic adaptation using stability measures
Xin Lei,Petar Aleksic +1 more
TL;DR: In this article, the authors describe a system for real-time acoustic adaptation using stability measures on a computer storage medium using a speaker adaptation profile and a stability measure for a segment of the transcription and determining that the stability measure satisfies a threshold.
Patent
Wake word evaluation
Stan Weidner Salvador,Jeffrey Paul Lilly,Frederick V. Weber,Jeffrey Penrod Adams,Ryan Paul Thomas +4 more
TL;DR: In this paper, a candidate word for evaluation as a wake word that activates a natural language control functionality of a computing device is provided, which may include one or more words or sounds.
Patent
Method and apparatus for automatically determining speaker characteristics for speech-directed advertising or other enhancement of speech-controlled devices or services
Harry Printz,Vikas Gulati +1 more
TL;DR: In addition to the primary information, human speech also conveys information concerning the speaker's gender, age, socioeconomic status, accent, language spoken, emotional state, or other personal characteristics, referred to as secondary information as mentioned in this paper.
References
More filters
Patent
Client/server architecture for text-to-speech synthesis
TL;DR: In this paper, a client/server text-to-speech synthesis system and method is proposed, where the server stores large databases for pronunciation analysis, prosody generation, and acoustic unit selection corresponding to a normalized text, while the client performs computationally intensive decompression and concatenation of selected acoustic units to generate speech.
Patent
Confusable word detection in speech recognition
Michael D. Riley,David Bjorn Roe +1 more
TL;DR: In this paper, after a vocabulary word is input into the system, a first set of phonemes representative of the vocabulary word was determined, and the first set was compared with a second set representative of a second vocabulary word.
Patent
Predicting auditory confusions using a weighted Levinstein distance
Michael Sabourin,Marc A. Fabiani +1 more
TL;DR: In this article, a confusability cost associated with two phonemic transcriptions is calculated based on a weighting of the Levinstein distance between the transcription pair, which measures the likelihood that a human or machine hearing the first word will mistakenly hear the second word.
Patent
Adaptive speech recognition with selective input data to a speech classifier
Lin Cong,Safdar M. Asghar +1 more
TL;DR: In this paper, a fuzzy Viterbi algorithm is used by a processor to compute maximum likelihood probabilities PR(O|λj) for each vocabulary word and the fuzzy distance measures and maximum likelihood probability are mixed in a variety of ways to preferably optimize speech recognition accuracy and speech recognition speed performance.
Patent
System and method for preventing enrollment of confusable patterns in a reference database
Marc A. Gardner,Tuan K. Nguyen +1 more
TL;DR: In this paper, a system and method for increasing the reliability of data matches between data stored in a database and newly provided data is proposed, which relies on the prescreening of newly-provided data patterns to insure that any newly provided dataset is not ambiguous with existing data patterns.