EP0784840A1

EP0784840A1 - Reading tutorial system

Info

Publication number: EP0784840A1
Application number: EP95933720A
Authority: EP
Inventors: Ellen Rubin Lebowitz; Ofer Bergman
Original assignee: RUBIN LEBOWITZ ELLEN STORYTIME Ltd; Bergman Ofer; Rubin Lebowitz Ellen
Current assignee: STORYTIME Ltd
Priority date: 1994-09-05
Filing date: 1995-09-05
Publication date: 1997-07-23
Also published as: WO1996007999A1; AU3625795A; IL110883A; IL110883A0; CA2199245A1

Abstract

A system including a text memory (24) having stored therein digital information representing a given reading text having indices at a plurality of text-locations, a sound memory (22) having stored therein digital information representing a prerecorded speech corresponding to the given text and having indices at a plurality of speech-locations corresponding, respectively, to the plurality of text-locations, a main processor (26) associated with the sound memory and the text memory which correlates between the speech-indices and the text-indices such that each text-location and its respective speech-location are substantially simultaneously addressable, a sound processor (32) associated with the main processor which processes digital information from the sound memory and provides an output corresponding to a reproduction of the prerecorded speech, a sound producing unit (36) which plays-back the reproduced speech to a user and a rate controller (30) associated with the sound processor which controls the rate at which the speech is reproduced, wherein the sound processor maintains the pitch of the reproduced speech substantially the same as the pitch of the prerecorded speech.

Description

1 READING TUTORIAL SYSTEM

2 FIELD OF THE INVENTION

3 The present invention relates to reading aids in

4 general and, more particularly, to devices and methods for

5 playing back sound information corresponding to a written

6 text.

7 BACKGROUND OF THE INVENTION

8 Reading accompanied by audible speech corresponding

9 to a text being read is known to be helpful in developing

10 reading skills, particularly for elementary school children

11 reading for the first time and for children having reading

12 disabilities such as dyslexia.

13 It is appreciated that speech accompaniment is

14 helpful to the reader, since it helps the reader associate

15 the graphemes he reads with their corresponding phonemes.

16 However, it is inconvenient, expensive and often

17 impossible to provide a child with a personal tutor who will

18 read texts aloud to the child. Therefore, the desired speech

19 information is normally prerecorded on a magnetic tape or the

20 like and played back while the child is reading the

21 corresponding text. Alternatively, the speech may be

22 digitally recorded on a computer memory.

23 A fundamental problem of this method, however, is

24 . that the rate of the played-back speech is rarely consistent

25 with the child's reading rate and, therefore, the above

26 mentioned association between graphemes and phonemes is

27 impaired.

28 Problems in dealing with reading disabilities are

29 outlined, for example, in "Development of Skill in Reading-

30 while-Listening", by "Margaret L. MacMahon", a Paper

31 presented at the 25th Annual Meeting of the International

32 Reading Association, St. Louis, Missouri, between May 5 and

33 May 9, 1980. The article describes experiments in which

34 speech was played-back at various rates to accompany reading -35 by children of different ages.

36 The results of these and other experiments indicate

37 that at fast play-back rates children have severe difficulty

38 in following the text, while it is believed that at very slow 1 play-back rates children tend to be bored with the reading.

2 Additionally, very slow play-back rates are expected to cause

3 inefficient reading habits. It is also appreciated that

4 reading rates vary considerably between children, even within

5 the same age group and, thus, there is no fixed rate which is

6 suitable for every child in a given group. Moreover, even if

7 on the average the speech rate is adapted to the reading rate

8 of a given child, it is not adapted to fluctuations in the

9 child's reading rate, for example due to difficulty in

10 reading certain words and phrases.

11 Playing back of prerecorded audible sounds, such as

12 speech, at a rate different from the original recording rate

13 is known in the art. When the recorded information is simply

14 played-back at a rate faster or slower than the original

15 recording rate, the pitch of the played-back sounds is higher

16 or lower than the original pitch. When the difference in

17 pitch is substantial, the reproduced audible sounds are

18 unpleasant, annoying and sometimes illegible. To overcome

19 this problem, a compensation in pitch is required.

20 U.K. Patent No. 2,229,068 describes a system for

21 playing back prerecorded audible sounds at a rate faster than

22 the original recording rate. The described system provides

23 pitch reduction compensation to maintain the played-back

24 sound substantially at the pitch of the prerecorded sound.

25 The sounds may be played-back at one of a number of discrete

26 rates, wherein appropriate pitch reduction is provided at

27 each rate. 28

29 30 31 32 33 34 -35 36 37 38 SUMMARY OF THE INVENTION The present invention seeks to provide a reading tutorial system which plays-back to a user at a controllable rate prerecorded sound information, preferably speech, while a corresponding reading text is being read by the user. According to the present invention, the prerecorded sound information is played back at a controllable rate, preferably at a rate adapted for the reading rate of the user. Adaptation of the play-back rate to the user's reading rate enables the user to more efficiently associate phonemes of the speech with corresponding graphemes of the text. Regular use of the present tutorial system is expected, in the long run, to improve reading skills of the user, such as phonological awareness and grapheme to phoneme translation ability. In a preferred embodiment of the present invention, information representing the reading text is stored in a text memory, while information representing the corresponding speech is stored in a sound memory. The text and sound memories are preferably both read-only computer memories and the memories are preferably both indexed in accordance with a preselected. indexing scheme. According to the present invention, the sound and text memories are correlated such that reference by the user to a given location in one of the memories is accompanied by automatic reference of the system to the corresponding location in the other memory. Thus, for example, when the user selects a location in the text where reading is to begin, the system plays-back the accompanying speech starting from a speech-location corresponding to the selected text-location. Additionally, the correlation between the sound memory and the text memory enables on-line indication of the text location corresponding to the speech location being played-back. According to one aspect of the present invention, the play-back rate is controlled by the user, preferably using a rate control member, such that the sound information is played-back substantially in accordance with the reading rate

- 3 -

BSTITUTESHEET KUL£2δ) of the user. According to another aspect of the present invention, the play-back rate is automatically controlled based on predetermined criteria, preferably criteria related to the actual reading rate. There is thus provided in accordance with a preferred embodiment of the invention, a reading tutorial system including: a text memory having stored therein digital information representing a given reading text having indices at a plurality of text-locations; a sound memory having stored therein digital information representing a prerecorded speech corresponding to the given text and having indices at a plurality of speech-locations corresponding, respectively, to the plurality of text-locations; a main processor associated with the sound memory and the text memory which correlates between the speech-indices and the text-indices such that each text-location and its respective speech-location are substantially simultaneously addressable; a sound processor associated with the main processor which processes digital information from the sound memory and provides an output corresponding to a reproduction of the prerecorded speech; a sound producing unit which plays-back the reproduced speech to a user; and a rate controller associated with the sound processor which controls the rate at which the speech is reproduced, wherein the sound processor maintains the pitch of the reproduced speech substantially the same as the pitch of the prerecorded speech. According to one preferred embodiment of the invention, the rate controller is controlled manually by the user to provide a desired play-back rate. The play-back rate may be selected from a plurality of discrete rates or the play-back rate may be continuously selectable. According to another preferred embodiment of the invention, the rate controller includes eye-tracking apparatus which determines the actual reading rate of the user and wherein the play-back rate is automatically adapted to the actual reading rate . In a preferred embodiment , the system further includes a display for displaying the reading text to the user . The display preferably includes a visual indicator which indicates to the user the text-location corresponding to a speech-location currently being played-back . In a preferred embodiment , the sound processor includes a digital signal processor . According to a preferred embodiment of the invention, for a given played-back speech rate , the processing rate of the sound processor varies in accordance with predetermined criteria dependent on characteristics of the prerecorded speech . Preferably, for played-back speech rates higher than the prerecorded speech rate , information representing consonants is processed at a rate lower than the processing rate of information representing vowels , and, for played-back speech rates lower than the prerecorded speech rate , information representing consonants is processed at a rate higher than the processing rate of information representing vowels . In accordance with an alternative, preferred, embodiment of the present invention, there is provided a reading tutorial system including: a text memory having stored therein digital information representing a given reading text having indices at a plurality of text-locations; a sound memory including a plurality of speech files, each speech file having stored therein digital information representing a digital reproduction of a prerecorded speech corresponding to the given text and having indices at a plurality of speech-locations corresponding, respectively, to the plurality of text-locations; a main processor associated with the sound memory and the text memory which correlates between the speech-indices and the text-indices such that each text-location and its respective speech-location in any of the speech files are

- 5 - substantially simultaneously addressable; a rate selector associated with the sound processor which selects the speech file from which the reproduced speech is to be played back; and a sound producing unit which plays-back the reproduced speech to a user, wherein each of the speech files defines a different, predetermined, reproduced speech rate. According to one variation of this embodiment of the invention, each speech file is a preprocessed speech file containing a digital reproduction of the prerecorded speech at a different, predetermined, respective, reproduced speech rate but at substantially the same pitch, and wherein all the speech files are reproduced from the same prerecorded speech. According to another variation of this embodiment of the invention, each speech file contains a digital reproduction of a different, respective, prerecorded speech having a predetermined, respective, prerecorded speech rate. In either of the above variations, the rate selector is preferably controlled manually by the user to provide a desired reproduced speech rate. In one preferred embodiment of the present invention, the sound memory and the text memory are both contained in a single read-only-memory (ROM) unit. Preferably, the ROM unit includes a CD-ROM unit. The CD-ROM unit preferably includes an optical disc. In another preferred embodiment of the invention, the sound memory and the text memory are both contained in a multi-user accessible memory unit. Further, in accordance with a preferred embodiment of the present invention there is provided a method for assisting a user in reading a given reading text including the steps of: storing digital information representing the given reading text indexed at a plurality of text-locations; storing digital information representing a prerecorded speech corresponding to the given text with indices at a plurality of speech-locations corresponding, respectively, to the plurality of text-locations; correlating between the speech-indices and the text- indices such that each text-location and its respective speech-location are substantially simultaneously addressable; processing digital information from the sound memory and providing an output corresponding to a reproduction of the prerecorded speech; playing-back the reproduced speech to the user; controlling the rate at which the speech is reproduced; and maintaining the pitch of the reproduced speech substantially the same as the pitch of the prerecorded speech. According to one preferred embodiment of the invention, the step of controlling the play-back rate includes the step of manually controlling the play-back rate. Preferably, the step of manually controlling the play-back rate includes the step of selecting the play-back rate from a plurality of discrete rates. Alternatively, the play-back rate is continuously selectable. According to another preferred embodiment of the invention, the step of controlling the play-back rate includes the steps of determining the actual reading rate of the user and automatically adapting the play-back rate to the actual reading rate. Preferably, the step of determining the actual reading rate includes the step of tracking the eye movement of the user. In a preferred embodiment of the invention, the method further includes the step of displaying the reading text to the user. Preferably, the step of displaying includes the step of visually indicating to the user the text-location corresponding to a speech-location currently being played- back. The method of the present invention may be used for teaching reading, for assisting reading of users having an eyesight disability, for assisting the reading of users having a reading disability and for teaching languages. In a preferred embodiment, the method further includes the step of supervising the user by determining whether the user follows the text and the speech. The step of supervising preferably includes the steps of introducing occasional inconsistencies between the text and the speech and determining whether the inconsistencies are detected by the 'user. In an additionally preferred embodiment of the present invention, the step of playing-back the reproduced speech includes the step of playing-back the reproduced speech at a predetermined volume level which excites the user phonologically and semantically. In a preferred embodiment of the invention, the step of correlating between the speech-indices and the text- indices includes the step of addressing a speech-location corresponding to a text-location selected by the user. Additionally or alternatively, in a preferred embodiment, the step of correlating between the speech-indices and the text- indices includes the step of addressing a text-location corresponding to a given speech-location. In accordance with an alternative embodiment of the present invention, there is provided a method for assisting a user in reading a given reading text including the steps of: storing digital information representing the given reading text indexed at a plurality of text-locations; storing a plurality of speech files, each speech file containing digital information representing a reproduction of a prerecorded speech corresponding to the given text and each speech file having indices at a plurality of speech-locations corresponding, respectively, to the plurality of text- locations; correlating between the speech-indices and the text- indices such that each text-location and its respective speech-location in any of the speech files are substantially simultaneously addressable; selecting the speech f ile from which the reproduced speech is to be played back; and playing-back the reproduced speech to the user, wherein each speech f ile def ines a di f ferent , respective, reproduced speech rate. One variation of this embodiment of the invention, further includes, in order to create each speech file, the step of preprocessing the prerecorded speech at a different, predetermined, respective, reproduced speech rate but at substantially the same pitch, wherein all the speech files are preprocessed from the same prerecorded speech. Another variation of this embodiment of the invention further includes, in order to create each of the speech files, the step of digitally reproducing a different, respective, prerecorded speech having a predetermined, respective, prerecorded speech rate. In accordance with a further, preferred, embodiment of the present invention, there is provided a read-only- memory (ROM) including: a text memory having stored therein digital information representing a given reading text having indices at a plurality of text-locations; and a sound memory having stored therein digital information representing a prerecorded speech corresponding to the given text and having indices at a plurality of speech-locations corresponding, respectively, to the plurality of text-locations. In accordance with another, preferred, embodiment of the present invention, there is provided a read-only-memory (ROM) including: a text memory having stored therein digital information representing a given reading text having indices at a plurality of text-locations; and a sound memory including a plurality of speech files, each speech file having stored therein digital information representing a digital reproduction of a prerecorded speech corresponding to the given text and having indices at a plurality of speech-locations corresponding, respectively, to the plurality of text-locations. In a preferred embodiment of the present invention, the ROM includes a CD-ROM. Preferably, the CD-ROM includes an optical disc. BRIEF DESCRIPTION OF THE DRAWINGS The present invention will be better understood from the following detailed description of preferred embodiments of the invention, taken in conjunction with the following drawings in which: ^* Fig. 1 is a simplified, pictorial, illustration of a reading tutorial system constructed and operative in accordance with a preferred embodiment of the present invention; Fig. 2 is a schematic block diagram functionally illustrating the system of Fig. 1; and Fig. 3 is a simplified, pictorial, illustration of a reading tutorial system constructed and operative in accordance with an alternative, preferred, embodiment of the present invention. DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT Reference is now made to Fig. 1, which schematically illustrates a reading tutorial system in accordance with a preferred embodiment of the present invention. The system preferably includes a central processing unit (CPU) 10 associated with a display 12 and a mouse 14 as known in the art. In a preferred embodiment of the present invention, the system further includes a sound producing device associated with CPU 10 and preferably including a head-set 20 adapted for a user 18. According to one embodiment of the present invention the system also includes a rate control pedal 16 operated by user 18 as described below. During operation of the system, user 18 reads a preselected text which is preferably displayed on display 12. A curser or other movable visual indicator, which may be controlled by mouse 14 or using a keyboard as known in the art, is preferably displayed together with the text on display 12. According to the present invention, a speech corresponding to the text being read by user 18 are played- back to the user via head-set 20 at a rate controlled by user 18 using rate controller 16. In a preferred embodiment of the invention, the curser or other visual indication moves along the text on display 12 according to the rate of the played-

- 10 -

SϋBSπTU7ESHEET RUL£ back speech. Reference is now made to Fig. 2 which functionally illustrates the system of Fig. 1. As shown in Fig. 2, the system preferably includes a sound memory 22 and a text memory 24, both of which are associated with a central processing unit (CPU) 26 which addresses the information stored in the memories. Memories 22 and 24 may be physically embodied in two, separate, digital memory units or in a single memory unit, as known in the art. Since the information stored in memories 22 and 24 is preferably fixed, read-only-memories (ROM) are preferably used, inter alia, to prevent user 18 from changing the stored information intentionally or accidentally. CPU 26 is preferably associated with a visual display 34 which displays the processed reading text and, via a digital signal processor (DSP) 32 and a digital-to-analog (D/A) converter 33, with a sound producing unit 36 which generates an audible reproduction of the prerecorded speech. Sound producing unit 36 is preferably associated with head-set 20 of user 18. Display 34 preferably includes a computer screen as indicated by reference numeral 12 in Fig. 1. Text memory 24 is used for storing digital information representing a given reading text, such as the content of a book, an essay or a reading exercise. According to a preferred embodiment of the invention, the text stored in memory 24 is indexed at preselected locations so as to enable access by CPU 26 to given locations of the text stored in memory 24. Any suitable indexing scheme may be used, for example indices may be provided at the beginning of each letter, syllable, word or sentence, so as to achieve a predetermined resolution in accessing the stored text. In accordance with a preferred embodiment of the present invention, sound memory 22 is used for storing digital information representing a prerecorded speech corresponding to the text stored in text memory 24. The speech stored in memory 22 is preferably indexed in accordance with the indexing scheme used for the text in memory 24. For example, if the information in memory 24 is indexed at the beginning of each word of the text, the information in memory 22 is preferably indexed at the beginning pf each, respective, Word of the corresponding prerecorded speech. As mentioned above, memories 22 and 24 may be embodied in separate memory units or both memories may be included in a single memory unit, preferably a read-only- memory (ROM) unit. In accordance with one preferred embodiment of the present invention, the speech information of memory 22 and the text information of memory 24 are both stored on a single CD-ROM unit, preferably including a compact optical disc. It should be appreciated that such CD- ROM units are capable of storing large volumes of speech and text information. The speech and text information stored on the CD-ROM unit is preferably indexed as described above. In accordance with another preferred embodiment of the present invention, the speech information of memory 22 and the text information of memory 24 are part of a central memory unit, such as a data-base. In this preferred embodiment of the invention, the speech and text information may be retrieved from the central memory unit by multiple users, using any known computer communication system or network. For example, the speech and text information may be stored in a data-base connected to InterNet. During operation, CPU 26 reads text information from memory 24 and corresponding speech information from memory 22. Pointer circuitry in CPU 26 correlates between the indices of the text information and the corresponding indices of the sound information, such that respective indices of memories 22 and 24 may be addressed simultaneously. When CPU 26 is directed by user 18 to address a desired location in the text, as described below, the above mentioned pointer circuitry also addresses the corresponding location in the speech to be played-back. As further shown in Fig. 2, CPU 26 is associated with a rate controller 30 which may be foot-operated, as shown by reference numeral 16 in Fig. 1, or hand-operated, for example, through appropriately defined functions of mouse 14 (Fig. 1). A analog-to-digital (A/D) converter is preferably employed to convert the generally analog output of rate controller .30 to a corresponding digital output readable by CPU 26. In accordance with the present invention, CPU 26 controls the rate of data processing by DSP 32 based on the input from rate controller 30. For example, in the embodiment of Fig. 1, the position of pedal 16 controls the output of the pedal and, thus, controls the processing rate of sound- bearing data by DSP 32. As known in the art, DSP 32 processes the sound- bearing digital data and D/A 33 generates a corresponding analog output to sound producing unit 36. The circuitry of sound producing unit 36 may include amplifiers, filters, etc., as required for reproducing the prerecorded speech through speakers (not shown) and/or head-set 20 (Fig. 1). It should be appreciated that the play-back rate of the reproduced speech is determined by the rate at which soundr bearing data is processed by DSP 32 and, therefore, the play- back rate is controlled by user 18 using rate controller 30. The data output rate of DSP 32 varies in accordance with the desired play-back rate, such that the data output rate is higher for higher play-back rates and lower for lower play-back rates. Alternatively, if DSP 32 is designed to output digital information at a given rate, down-sampling of the sound-bearing digital data, i.e. processing of only part of the digital data, may be used for play-back rates higher than the original speech rate. To maintain the desired data output rate for play- back rates lower than the original speech rate, DSP 32 preferably up-sa ples the sound-bearing digital data, i.e. generates additional samples which may be duplicates of adjacent existing samples or otherwise dependent on existing samples. If up-sampling is not used, the data output rate of DSP 32 varies in accordance with the desired play-back rate. It is appreciated that, in natural speech, changes in rate may be inhomogeneous, e.g. the time-span of vowels is generally more dependent on the speech rate than the time span of consonants. Thus, in a preferred embodiment of the

- 13 - present invention, changes in played-back speech rate are not homogeneous. For example, changes in the processing rate of data strings representing consonants may be different from, and generally proportionally lower than, changes in the processing rate of data strings representing vowels. To distinguish between different speech elements, such as vowels and consonants, the corresponding data-strings may be marked to indicate the appropriate changes in processing rate required for each data-string. It is appreciated, however, that the pitch of the played-back speech varies with the play-back rate, i.e. the higher the play-back rate the higher the pitch. Thus, according to the present invention, the pitch of the played- back speech is controlled in accordance with the play-back rate. The pitch is preferably controlled by pitch compensation circuitry which receives from CPU 26 a pitch control input responsive to the play-back rate and provides appropriate pitch compensation. Since the required change in pitch is uniquely determined by the change in play-back-rate, pitch compensation may be based on a predetermined formula executed by the pitch compensation circuitry, as known in the art. The pitch compensation circuitry may be included in DSP 32, as shown in Fig. 2, or it may be provided in a separate unit preceding or following DSP 32. A preferred sequence of operation of the present tutorial system will now be described, referring also to Fig. 1. User 18 uses mouse 14 to select a preselected portion of the reading text to appear on display 34. The exact location from which reading is to begin is preferably highlighted or otherwise distinguished on display 34 as known in the art. The pointer circuitry of CPU 26 identifies the index of the selected location in text memory 24 and addresses the corresponding index in speech memory 22. Thus, the prerecorded speech is played-back starting from the location selected by user 18. In a preferred embodiment of the invention, the highlighted location in the displayed text, which may be a letter, a syllable, a word, etc., moves in accordance with the play-back rate of the corresponding

- 14 - speech. It should be appreciated that due to the indexing scheme which correlates between memories 22 and 24, user 18 can use mouse 14 to "hop" to any desired location in the text, preceding or succeeding the initial location, while listening to the corresponding speech location after each "hop". If the initial play-back rate is unsuitable for user 18, i.e. too fast or to slow, user 18 changes the play back rate using rate controller 30. The pitch of the played-back speech is preferably substantially constant, due to the automatic pitch compensation described above. This, preferably on-line, control of the play-back rate ensures that the prerecorded speech is played-back to the user at a rate adapted for his or her specific reading skills and/or habits. In a preferred embodiment of the invention, rate controller 30 provides continuous rate control. However, in a simpler system, controller 30 may be embodied as a multi- position switch, wherein a plurality of discrete play-back rates are defined by the different switch positions. Alternatively, the play-back rate may be selected from a menu appearing on display 34 using a keyboard (not shown) or mouse 14. In an alternative, preferred, embodiment of the present invention, changes in the speech rate and appropriate pitch compensations are performed off-line rather than on- line. According to this embodiment, preprocessed files corresponding to a plurality of different play-back rates of the prerecorded speech are stored separately in speech memory 22. To correlate between the preprocessed files and the text in memory 24, each of the preprocessed files is preferably indexed in accordance with the indexing scheme of text memory 24. At any given time during operation of the system, sound information is retrieved from one of the preprocessed files which corresponds to the play-back rate selected by user 18 from a preselected menu, for example by using a multi-position switch as described above. In a

- 15 -

SUBSTITUTE SHEET (HU E 2 ) preferred embodiment of the invention, the speech location being played-back is substantially unaffected by changes in the play-back rate, as described below. Since each preprocessed file preferably corresponds to a constant play-back rate, having a predetermined constant ratio relative to the original, prerecorded, speech rate, the ratios between the play-back rates of the different preprocessed files are also constant and predeterminable. Thus, based on a given speech location in a given preprocessed file, it is possible to accurately determine a corresponding speech location in any of the other preprocessed files. For example, if the ratio between the play-back rates of two preprocessed files is 2:1, there is a time ratio of 2:1 between corresponding speech locations of the two files. This time ratio is preferably applied to maintain a correct speech location when the user switches between the different play-back rates. Alternatively, since the same indexing scheme is preferably used for all of the preprocessed files, the indexing of the preprocessed files can be utilized to maintain the correct speech location when the play-back rate is changed by user 18. To obtain the preprocessed files, processing as described above is preferably employed to change the speech rate and to provide appropriate pitch compensation for each file. For example, down-sampling or up-sampling as described above can be used. The preprocessed files are then stored separately in speech memory 22. Therefore, no further processing is required, on-line, to provide the desired play- back rate and appropriate pitch compensation during operation. In a further, preferred, embodiment of the invention, a plurality of prerecordings of the original speech are used for providing the different speech rates, whereby the text is read at a different, preselected, speech rate during each prerecording. The prerecorded speeches are then stored separately in speech memory 22, e.g. in separate files. The provision of a plurality of preselected actual speech rates, at the prerecording stage, obviates the need for processing as in the above embodiments to provide different play-back rates. During operation, the prerecorded speeches are retrieved from speech memory 22, in accordance with the rate- selections of user 18, as described above with reference to the embodiment in which preprocessed files are used. The same indexing scheme is preferably used for all the prerecorded speeches so as to maintain the correct speech location when user 18 switches between different speech rates. According to another preferred embodiment of the present invention, not shown in the drawings, the play back rate is controlled automatically using an eye-tracking system. For example, as shown in Fig. 3, the eye tracking system may include an optical sensor 40, such as a video camera, which follows the movement of the pupils of user 18. The output of optical sensor 40 is preferably processed by appropriate rate-control circuitry in controller 30 or CPU 26. According to this embodiment of the invention, user 18 simply reads the text while the accompanying speech is automatically played-back at the actual reading rate of the user. Eye-tracking devices as required for this preferred embodiment of the present invention are known in the art. According to a further, preferred, embodiment of the present invention, the reading tutorial system provides means for supervising the user by determining whether the user follows the text and the speech with sufficient concentration. This can be achieved, for example, by introducing occasional inconsistencies between the text and the speech, whereby the user is required to provide a preselected active response each time an inconsistency is detected. Reference is again made to Fig. 2. In a preferred embodiment of the present invention, the sounds produced by sound producing unit 36 are volume-controlled, for example by an appropriate control button on unit 36. According to this preferred embodiment, the speech accompaniment can be played- back at very low volume levels so as to cause subliminal phonological and semantic excitation of the user, as known in 1 the art. With such low volume speech accompaniment, the user

2 is prevented from being fully dependent on the played-back

3 speech during reading.

4 It should be appreciated that use of the present

5 reading tutorial system is not limited to improvement of

6 reading skills among school children. In fact, the present

7 invention may be equally suitable for a variety of other

8 uses, for example learning of new languages, assisting

9 reading of people having poor eyesight and/or reading

10 disabilities such as dyslexia.

11 Speech memory 22 and text memory 24 may be

12 implemented form of memory known in the art, such information

13 may be stored using any suitable memory. For example, both

14 the text and sound memory

15 It will be appreciated by persons skilled in the art

16 that the present invention is not limited to what has been

17 thus far described. Rather, the scope of the present

18 invention is limited only by the following claims: 19

20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 ^"35 36 37 38

Claims

1 C L A I M S 2

3 1. A reading tutorial system comprising:

4 a text memory having stored therein digital

5 information representing a given reading text having indices

6 at a plurality of text-locations;

7 a sound memory having stored therein digital

8 information representing a prerecorded speech corresponding

9 to the given text and having indices at a plurality of

10 speech-locations corresponding, respectively, to the

11 plurality of text-locations;

12 a main processor associated with the sound memory and

13 the text memory which correlates between the speech-indices

14 and the text-indices such that each text-location and its

15 respective speech-location are substantially simultaneously

16 addressable;

17 a sound processor associated with the main processor

18 which processes digital information from the sound memory and

19 provides an output corresponding to a reproduction of the

20 prerecorded speech;

21 a sound producing unit which plays-back the

22 reproduced speech to a user; and

23 a rate controller associated with the sound processor

24 which controls the rate at which the speech is reproduced,

25 wherein the sound processor maintains the pitch of

26 the reproduced speech substantially the same as the pitch of

27 the prerecorded speech. 28

29 2. A system according to claim 1 wherein the rate

30 controller is controlled manually by the user to provide a

31 desired play-back rate. 32

33 3. A system according to claim 2 wherein the play-back

34 rate is selected from a plurality of discrete rates. -35

36 4. A system according to claim 2 wherein the play-back

37 rate is continuously selectable.

38

19 - 1 5. A system according to claim 1 wherein the rate

2 controller comprises eye-tracking apparatus which determines

3 the actual reading rate of the user and wherein the play-back

4 rate is automatically adapted to the actual reading rate. 5

6 6. A system according to any of the preceding claims

7 wherein the sound processor comprises a digital signal

8 processor. 9

10 7. A system according to any of claims 1-5 wherein, for

11 a given played-back speech rate, the processing rate of the

12 sound processor varies in accordance with predetermined

13 criteria dependent on characteristics of the prerecorded

14 speech. 15

16 8. A system according to claim 7 wherein, for played-

17 back speech rates higher than the prerecorded speech rate,

18 information representing consonants is processed at a rate

19 lower than the processing rate of information representing

20 vowels, and, for played-back speech rates lower than the

21 prerecorded speech rate, information representing consonants

22 is processed at a rate higher than the processing rate of

23 information representing vowels. 24

25 9. A reading tutorial system comprising:

26 a text memory having stored therein digital

27 information representing a given reading text having indices

28 at a plurality of text-locations;

29 a sound memory including a plurality of speech files,

30 each speech file having stored therein digital information

31 representing a digital reproduction of a prerecorded speech

32 corresponding to the given text and having indices at a

33 plurality of speech-locations corresponding, respectively, to

34 the plurality of text-locations;

"35 a main processor associated with the sound memory and

36 the text memory which correlates between the speech-indices

37 and the text-indices such that each text-location and its

38 respective speech-location in any of the speech files are substantially simultaneously addressable; a rate selector associated with the sound processor which selects the speech file from which the reproduced speech is to be played back; and a sound producing unit which plays-back the reproduced speech to a user, wherein each of the speech files defines a different, predetermined, reproduced speech rate .

10. A system according to claim 9 wherein at least one speech file is a preprocessed speech file containing a digital reproduction of the prerecorded speech at a different, predetermined, respective, reproduced speech rate but at substantially the same pitch , and wherein all the speech files are reproduced from the same prerecorded speech :

11 . A system according to claim 9 wherein at least one speech file contains a digital reproduction of a different , respective , prerecorded speech having a predetermined , respective, prerecorded speech rate.

12. A system according to any of claims 9-11 wherein the" rate selector is controlled manually by the user to provide a desired reproduced speech rate.

13. A system according to any of claims 9-11 and further comprising a display for displaying the reading text to the user .

14. A system according to claim 13 wherein the display comprises a visual indicator which indicates to the user the text-location corresponding to a speech-location currently being played-back.

15. A system according to any of claims 1-5 or 9-11 wherein said sound memory and said text memory are both contained in a single read-only-memory (ROM) unit.

- 21 - 1 16. A system according to claim 15 wherein the ROM unit

2 comprises a CD-ROM unit. 3

4 17. A system according to claim 16 wherein the CD-ROM

5 unit comprises an optical disc. 6

7 18. A system according to any of claims 1-5 or 9-11

8 wherein said sound memory and said text memory are both

9 contained in a multi-user accessible memory unit. 10

11 19. A method for assisting a user in reading a given

12 reading text comprising the steps of:

13 storing digital information representing the given

14 reading text indexed at a plurality of text-locations;

15 storing digital information representing a

16 prerecorded speech corresponding to the given text with

17 indices at a plurality of speech-locations corresponding,

18 respectively, to the plurality of text-locations;

19 correlating between the speech-indices and the text-

20 indices such that each text-location and its respective

21 speech-location are substantially simultaneously addressable;

22 processing digital information from the sound memory^'

23 and providing an output corresponding to a reproduction of

24 the prerecorded speech;

25 playing-back the reproduced speech to the user;

26 controlling the rate at which the speech is

27 reproduced; and

28 maintaining the pitch of the reproduced speech

29 substantially the same as the pitch of the prerecorded

30 speech. 31

32 20. A method according to claim 19 wherein the step of

33 controlling the play-back rate comprises the step of manually

34 controlling the play-back rate. ^"35

36 21. A method according to claim 20 wherein the step of

37 manually controlling the play-back rate comprises the step of

38 selecting the play-back rate from a plurality of discrete

- 22 -

SUBSTITUTE SHEET (RULE 2 1 rates . 2

3 22. A method according to claim 21 wherein the play-back

4 rate is continuously selectable. 5

6 23. A method according to claim 19 wherein the step of

7 controlling the play-back rate comprises the steps of

8 determining the actual reading rate of the user and

9 automatically adapting the play-back rate to the actual 10 reading rate.

11

12 24. A method according to claim 23 wherein the step of

13 determining the actual reading rate comprises the step of

14 tracking the eye movement of the user. 15

16 25. A method for assisting a user in reading a given

17 reading text comprising the steps of:

18 storing digital information representing the given

19 reading text indexed at a plurality of text-locations;

20 storing a plurality of speech files, each speech file

21 containing digital information representing a reproduction of

22 a prerecorded speech corresponding to the given text and each

23 speech file having indices at a plurality of speech-locations

24 corresponding, respectively, to the plurality of text-

25 locations;

26 correlating between the speech-indices and the text-

27 indices such that each text-location and its respective

28 speech-location in any of the speech files are substantially

29 simultaneously addressable;

30 selecting the speech file from which the reproduced

31 speech is to be played back; and

32 playing-back the reproduced speech to the user,

33 wherein each speech file defines a different,

34 respective, reproduced speech rate. -35

36 26. A method according to claim 25 and further

37 comprising, to create each speech file, the step of

38 preprocessing the prerecorded speech at a different, predetermined, respective, reproduced speech rate but at substantially the same pitch, wherein all the speech files are preprocessed from the same prerecorded speech.

27. A method according to claim 25 and further comprising, to create each of the speech files, the step of digitally reproducing a different, respective, prerecorded speech having a predetermined, respective, prerecorded speech rate.

28. A method according to any of claims 19 - 27 and further comprising the step of displaying the reading text to the user.

29. A method according to claim 28 wherein the step of displaying comprises the step of visually indicating to the user the text-location corresponding to a speech-location currently being played-back.

30. A method according to any of claims 19 - 27 used for teaching reading.

31. A method according to any of claims 19 - 27 used for assisting reading of users having an eyesight disability.

32. A method according to any of claims 19 - 27 used for assisting the reading of users having a reading disability.

33. A method according to any of claims 19 - 27 used for teaching a language.

34. A method according to any of claims 19 - 27 and further comprising the step of supervising the user by determining whether the user follows the text and the speech.

35. A method according to claim 34 wherein the step of supervising comprises the steps of introducing occasional inconsistencies between the text and the speech and determining whether the inconsistencies are detected by the user.

36. A method according to any of claims 19 - 27 wherein the step of playing-back the reproduced speech comprises the step of playing-back the reproduced speech at a predetermined volume level which excites the user phonologically and semantically.

37. A method according to any of claims 19 - 27 wherein the step of correlating between the speech-indices and the text-indices comprises the step of addressing a speech- location corresponding to a text-location selected by the user.

38. A method according to any of claims 19 - 27 wherein the step of correlating between the speech-indices and the text-indices comprises the step of addressing a text-location corresponding to a given speech-location.

39. A read-only-memory (ROM) comprising: a text memory having stored therein digital information representing a given reading text having indices at a plurality of text-locations; and a sound memory having stored therein digital information representing a prerecorded speech corresponding to the given text and having indices at a plurality of speech-locations corresponding, respectively, to the plurality of text-locations.

40. A read-only-memory (ROM) comprising: a text memory having stored therein digital information representing a given reading text having indices at a plurality of text-locations; and a sound memory including a plurality of speech files, each speech file having stored therein digital information representing a digital reproduction of a prerecorded speech corresponding to the given text and having indices at a plurality of speech-locations corresponding, respectively, to the plurality of text-locations.

41. A ROM according to claim 39 or claim 40 comprising a CD-ROM. i 42. A ROM according to claim 41 comprising an optical disc.