Skip to content Skip to navigation Skip to collection information

Connexions

You are here: Home » Content » ELEC 301 Project: Voice Recognition » 3 - Voice Recognition: Chosen Methods of Investigation

Navigation

Recently Viewed

This feature requires Javascript to be enabled.
 

3 - Voice Recognition: Chosen Methods of Investigation

Module by: Kevin Chu. E-mail the author

Chosen Methods of Investigation

The limited timeframe of our project meant both DTW and HMM-based approaches were impractical, requiring many hundreds more man-hours than was available. We chose to focus on achieving solid results from a more primitive algorithm, the LPC, and work on making it more robust thereafter.

We collected the several hundreds of data samples used to train the library from ourselves.

We featured-matched input and stored data using the Yule-Walker autocorrelation method, minimizing the forward prediction error in the least squares sense. This was done using Matlab’s Yule-Walker AR Estimator.

Testing the algorithm resulted in an abysmal 20-30% accuracy.

We thought to produce better base accuracy with an algorithm of our own making. Our final results are based upon the following algorithm outlined:

  1. Convolution-based segmentation
  2. Feature extraction of formants via nonlinear power filter
  3. Display filtered spectrum on a discrete, weighted scatter plot
  4. Trace out contours of the maximum-likelihood Gaussian Mixture Model (GMM) using a maximum-likelihood GMM estimator
  5. Construct a standardized GMM parameter library for each number
  6. Find the GMM matching the input with a maximum-likelihood fit

an envelope with a blue page

Collection Navigation

Content actions

Download:

Collection as:

PDF | EPUB (?)

What is an EPUB file?

EPUB is an electronic book format that can be read on a variety of mobile devices.

Downloading to a reading device

For detailed instructions on how to download this content's EPUB to your specific device, click the "(?)" link.

| More downloads ...

Module as:

PDF | More downloads ...

Add:

Collection to:

My Favorites (?)

'My Favorites' is a special kind of lens which you can use to bookmark modules and collections. 'My Favorites' can only be seen by you, and collections saved in 'My Favorites' can remember the last module you were on. You need an account to use 'My Favorites'.

| A lens I own (?)

Definition of a lens

Lenses

A lens is a custom view of the content in the repository. You can think of it as a fancy kind of list that will let you see content through the eyes of organizations and people you trust.

What is in a lens?

Lens makers point to materials (modules and collections), creating a guide that includes their own comments and descriptive tags about the content.

Who can create a lens?

Any individual member, a community, or a respected organization.

What are tags? tag icon

Tags are descriptors added by lens makers to help label content, attaching a vocabulary that is meaningful in the context of the lens.

| External bookmarks

Module to:

My Favorites (?)

'My Favorites' is a special kind of lens which you can use to bookmark modules and collections. 'My Favorites' can only be seen by you, and collections saved in 'My Favorites' can remember the last module you were on. You need an account to use 'My Favorites'.

| A lens I own (?)

Definition of a lens

Lenses

A lens is a custom view of the content in the repository. You can think of it as a fancy kind of list that will let you see content through the eyes of organizations and people you trust.

What is in a lens?

Lens makers point to materials (modules and collections), creating a guide that includes their own comments and descriptive tags about the content.

Who can create a lens?

Any individual member, a community, or a respected organization.

What are tags? tag icon

Tags are descriptors added by lens makers to help label content, attaching a vocabulary that is meaningful in the context of the lens.

| External bookmarks