Pre-Grant Publication Number: 20100262423
Feature compensation approach to robust speech recognition
Please help the USPTO examine the application by evaluating the relevance of the publicly submitted prior art to the patent application. Peer To Patent forwards the Top 10 most relevant prior art submissions and their annotations to the USPTO.

Review this prior art and click on the thumbs up (or down) to indicate whether this submission should be forwarded to the USPTO.
If you login then you can add an annotation by typing in the box at the bottom of the screen to comment on the relevance of the prior art to the claims of the patent application.
Prior Art Detail
Annotations(0)
#636The application of Taylor series
Applies to Claims 1,10,2,3,4,5,6,7,8,9
Submitted by: Mehmet YilmazLast updated: over 2 years ago
1 thumb up 0 thumbs down
Summary / Description
Summary / DescriptionIt appears the use of vector taylor series approximation in speech has already existed
 
Basic Information
Type of Prior ArtPrint Publication
 
Publication Title *OVERCOMING THE VECTOR TAYLOR SERIES APPROXIMATION IN SPEECH FEATURE ENHANCEMENT — A PARTICLE FILTER
 
AuthorFriedrich Faubel and Matthias Wolfel
 
ISBN
 
Page Range
 
MediumJournal article
 
Publication Date *January 1, 2007
 
URLhttp://citeseerx.ist.psu.edu/vi...
 
Notes / To Do
NotesThis paper provides evidence the community has explored and published methods of filtering speech based on the taylor series approximation and has already moved to supplemental methods.
 
Excerpt
Excerpt
We present a simple, fast and previously unreported noise compensation method for particle filter (PF) based speech feature enhancement, which outperforms the vector Taylor series noise compensation method used by current PF approaches in terms of speed as well as word error rate. Furthermore, we devise a fast acceptance test that overcomes the particle decimation problem associated with PFs for speech feature enhancement, which makes the particle filter approach computationally more efficient.
 
Relevance
Claims
1
In a computing environment, a method comprising, receiving feature vectors for an unknown utterance, compensating for additive noises or convolutional distortions, or both additive noise and convolutional distortions, including by using a high-order vector Taylor series approximation of a model of distortions to provide compensated feature vectors to a speech recognizer.
Relevance
Invalid due to existing research
Invalid due to existing research
Claim Chart
All
2
The method of Claim 1 wherein the feature vectors are cepstral domain feature vectors, and further comprising, using a plurality of frames to estimate noise model parameters in the cepstral domain.
Relevance
Tied to claim 1
Tied to claim 1
Claim Chart
All
3
The method of Claim 2 further comprising, transforming the noise model parameters from the cepstral domain to log power-spectral domain noise model parameters.
Relevance
Tied to claim 2 to 1
Tied to claim 2 to 1
Claim Chart
All
4
The method of Claim 3 further comprising, training with clean speech to produce at least one Gaussian mixture model used in transforming the noise model parameters.
Relevance
Tied to claim 3 to 2 to 1
Tied to claim 3 to 2 to 1
Claim Chart
All
5
The method of Claim 4 wherein training with clean speech further comprises performing maximum likelihood training to produce acoustic models.
Relevance
Tied to claim 4 to 3 to 2 to 1
Tied to claim 4 to 3 to 2 to 1
Claim Chart
All
6
The method of Claim 3 wherein using the high-order vector Taylor series approximation comprises computing relevant statistics representing the log power-spectral domain noise model parameters.
Relevance
Tied to claim 3 to 2 to 1
Tied to claim 3 to 2 to 1
Claim Chart
All
7
The method of Claim 6 further comprising, transforming the relevant statistics from the log power-spectral domain into transformed statistics in the cepstral domain.
Relevance
Tied to claim 6 to 3 to 2 to 1
Tied to claim 6 to 3 to 2 to 1
Claim Chart
All
8
The method of Claim 7 further comprising, using the transformed statistics to re-estimate the noise model parameters.
Relevance
Tied to claim 7 to 6 to 3 to 2 to 1
Tied to claim 7 to 6 to 3 to 2 to 1
Claim Chart
All
9
The method of Claim 8 further comprising, using the re-estimated noise model parameters to provide the compensated feature vectors to the speech recognizer.
Relevance
Tied to claim 8 to 7 to 6 to 3 to 2 to 1
Tied to claim 8 to 7 to 6 to 3 to 2 to 1
Claim Chart
All
10
The method of Claim 7 further comprising, normalizing the compensated feature vectors.
Relevance
Tied to claim 7 to 6 to 3 to 2 to 1
Tied to claim 7 to 6 to 3 to 2 to 1
Claim Chart
All
0 days left
Activity
Application
Discussion (0)
Prior Art (3)
Research (0)
Subscribe to this Community
Invite a Reviewer
HOW TO USE PEER TO PATENT
Peer To Patent Videos