Granted Patent B2

US 9,911,349 · App. 13/163,193 · Granted Mar 6, 2018

Inventors: Adithya Renduchintala (Boulder, CO); Robin Smith (Harrisonburg, VA)

Assignee: ROSETTA STONE, LTD.

G09B5/06G09B19/06

Abstract

Claims (35)

1. A method, comprising:

identifying, via a processor at a language instruction device, a speech segment in a target language that is susceptible to mispronunciation by language learners;

storing a data file including digital speech data representative of the speech segment;

selecting an auditory attribute for use in playing the identified speech segment by the language instruction device;

manipulating, via the processor, the data file to produce first manipulated digital speech data including a first altered level of the auditory attribute that differs from a naturally occurring level of the auditory attribute;

playing, via a speaker operatively coupled to the language instruction device, the first manipulated digital speech data including at least one instance of the identified speech segment having the first altered level of the auditory attribute;

receiving, from a language student and in response to the playing, a student utterance that includes the identified speech segment;

converting, via the processor, the student utterance to digital utterance data;

comparing, via the processor and in response to the converting, the digital utterance data with the digital speech data to define an error level of a pronunciation of the identified speech segment by the language student;

manipulating, via the processor and based on the error level, the data file to produce second manipulated digital speech data including a second altered level of the auditory attribute; and

playing, via the speaker, the second manipulated digital speech data including at least one instance of the identified speech segment having the second altered level of the auditory attribute.

2. The method of claim 1 , further comprising:

diminishing an extent of alteration of the first altered level of the auditory attribute of the first manipulated digital speech data played by the language instruction system as the error level of the pronunciation of the identified speech segment by the student declines.

3. The method of claim 1 , wherein the identified speech segment is at least one of: (a) a letter; (b) a phoneme; (c) a syllable; or (d) a word.

4. The method of claim 1 , wherein the auditory attribute includes at least one of: (a) pitch; (b) volume; (c) speed of play; or (d) tone.

5. The method of claim 1 , wherein the receiving the student utterance includes recording the student utterance by the language instruction device.

6. A method, comprising:

identifying, at a computer, a first speech segment and a second speech segment in a target language, the first speech segment and the second speech segment susceptible to being pronounced similarly by a language student;

storing a first data file including digital speech data representative of the first speech segment and a second data file including digital speech data representative of the second speech segment;

manipulating, at the computer, the first data file to produce first manipulated digital speech data including an increased pitch level of the first speech segment with respect to a normally occurring pitch level of the first speech segment;

manipulating, at the computer, the second data file to produce second manipulated digital speech data including a decreased pitch level of the second speech segment with respect to a normally occurring pitch level of the second speech segment;

playing, via a speaker operatively coupled to the computer, at least one instance of the first manipulated digital speech data and at least one instance of the second manipulated digital speech data;

receiving, from the language student and in response to the playing, a student utterance of the first speech segment and a student utterance of the second speech segment;

converting, at the computer, the student utterance of the first speech segment to first digital utterance data and the student utterance of the second speech segment to second digital utterance data;

comparing, at the computer, the first digital utterance data with the digital speech data representative of the first speech segment to define a first error level of the student utterance of the first speech segment;

comparing, at the computer, the second digital utterance data with the digital speech data representative of the second speech segment to define a second error level of the student utterance of the second speech segment; and

gradually reducing, at the computer, an amount of manipulation of the first data file and the second data file as the first error level and the second error level decline.

7. The method of claim 6 , wherein the first speech segment includes an utterance of the letter “L,” and the second speech segment includes an utterance of the letter “R”.

8. The method of claim 7 , further comprising:

receiving, at the computer, input from the language student, the input including an utterance of a second text sequence that includes at least one utterance of the letter “L” and at least one utterance of the letter “R”.

9. The method of claim 8 , further comprising:

measuring an error level of a pronunciation of the at least one utterance of the letter “L” and the at least one utterance of the letter “R”.

10. The method of claim 6 , wherein:

the manipulating the first data file includes adjusting a pitch level of an utterance of the letter “L” in the first file at the computer, in accordance with a measured error level of the pronunciation of the letter “L”; and

the manipulating the first data file includes adjusting a pitch level of an utterance of the letter “R” in the second data file at the computer, in accordance with a measured error level of the pronunciation of the letter “R”.

Assignments (12)

Continuity (1)