US3215821A

US3215821A - Speech-controlled apparatus and method for operating speech-controlled apparatus

Info

Publication number: US3215821A
Application number: US837220A
Authority: US
Inventors: Walter H Stenby
Original assignee: Individual
Current assignee: Individual
Priority date: 1959-08-31
Filing date: 1959-08-31
Publication date: 1965-11-02
Anticipated expiration: 1982-11-02

Description

Nov. 2, 1965 W. H. STENBY CONTROLLED AP SPEECH- PARATUS AND METHOD FOR OPERATING SPEECH-CONTROLLED APPARATUS Filed Aug. 31, 1959 IIIITI I ATTORNEY United States Patent 3,215,821 SPEECH-CONTROLLED APPARATUS AND METH- OD FOR SPERATING SPEECH-CQNTROLLED APPARATUS Walter H. Steuby, 98 Riverside Drive, New York, NY. Filed Aug. 31, 1959, Ser. No. 837,220 16 (Ilaims. (Cl. 235-151) The present invention relates to a method for operating apparatus under control of the human voice, and to appai ratus controlled by the human voice to perform operations, such as calculating operations.

Human speech consists of phonetic sounds which are basically divided into vowels and consonants. It is well known that the phonetic sounds are not identical with the letters of the alphabet, and that, for example, the English language comprises 36 phonetic sounds.

When the phonetic sounds are spoken into a microphone, each phonetic sound is converted into a complex electrical wave, and it is known to provide tuned means, which, in theory, are responsive only to the complex electrical wave representing a single phonetic sound. Such tuned means have been used for operating printing means representing the phonetic sounds, so that symbols representing such phonetic sounds can be printed as a speaker speaks into a microphone. For each pronounced sound, a phonetic symbol is printed.

The following U.S. patents disclose sound printing mechanism of this type: Dudley 2,195,081, Dreyfus 2,540,660, Kalfaian 2,613,273, Kalfaian 2,673,893 and Kalfaian 2,705,260.

However, due to the fact that different speakers pronounce theoretically identical phonetic sounds in a different manner, the apparatus of the prior art is not completely reliable to produce exact phonetic transcriptions of spoken words.

It is the principal object of the present invention to provide a speech controlled apparatus which performs different operations when different commands are spoken.

It is one object of the present invention to provide a new method for operating voice-controlled apparatus in a manner which is extremely reliable and eliminates errors caused by wrong identification of phonetic sounds by the apparatus.

Another object of the present invention is to provide a speech responsive apparatus responsive only to spoken selected sequences of selected phonetic sounds whose acoustic characteristics are substantially different so that the apparatus can reliably separate and distinguish the selected phonetic sounds from each other.

Another object of the present invention is to provide a method employing not all phonetic sounds of the English language, for controlling an apparatus, or of any other language, but using only a limited number of representative phonetic sounds which are selected from groups of similar phonetic sounds to be easily identified and dis tinguished by tuned means of the type disclosed in the above listed U.S. patents.

Another object of the present invention is to provide a method for controlling apparatus by speech wherein the speaker does not use any existent language, but sequences of a limited number of easily distinguishable phonetic sounds.

Another object of the present invention is to provide only short sequences of such selected distinguishable phonetic sounds, avoiding the longer words of existent languages.

Another object of the present invention is to simplify speech controlled apparatus by operating it by short sequences of selected easily distingiushable phonetic sounds for representing information or commands.

'ice

A further object of the present invention is to provide computing apparatus which can be controlled by speech to perform totalizing or other computing operations.

A further object of the present invention is to enable space travelers to operate machines by speech while confined in a narrow space.

With these objects in view one aspect of, the present invention relates to a method for operating apparatus by speaking. The method comprises the steps of sounding selected sequences of selected phonetically and acoustically easily distinguishable phonetic sounds, each of which represents a group of similar phonetic sounds; converting each pronounced sequence of phonetic sounds into an electric command signal; and supplying the command signals to apparatus adapted to perform different operations under control of command signals corresponding to different sounded sequences of phonetic sounds.

In accordance with the present invention, a limited number of phonetic sounds is used, and many of the 36 phonetic sounds of the English language, and of the additional sounds of other languages, are not used at all. In this manner, a limited number of phonetic sounds can be selected which are distinct, and easily distinguishable by tuned means of the type disclosed in the above listed patents. Other phonetic sounds which are frequently pronounced by different speakers in such a manner that they cannot be distinguished from each other are not used in the method of the present invention. For example, only one phonetic sound, namely s as in see is used of the group of similar sounds appearing underlined in the following words: see, crazy, there, thing, shot, leisure, chip, joy.

Furthermore, in accordance with the present invention the sequences of selected phonetic sounds do not necessarily represent words in any existent language. Short sequences of two to four phonetic sounds are preferred in accordance with the present invention so that the voice controlled apparatus can be considerably simplified if the method of the present invention is applied.

In other words, the operator does not control speech re sponsive apparatus by speaking English, or any other existent language, but by speaking an artificial language consisting of short words made up of distinct phonetic sounds having characteristic frequency components.

The phonetic sounds used in accordance with the present invention are so selected that they are of the type pronounced very similarly by different speakers, and even by speakers whose native languages are different, such as French and English. For example, the th sound is preferably omitted from the phonetic sounds used in accordance with the present invention, since it is rarely correctly pronounced by persons whose native language is not English.

The U.S. Patent 2,540,660 shows in FIG. 1 the frequency components of the main phonetic elements of French sounds. Similar graphs can be obtained for English phonetic sounds, and from such graphs, as well as by experimentation, see U.S. Patent 2,646,465, the limited number of phonetic sounds can be determined and selected which have easily distinguishable frequency components and characteristics. For example, it is proposed, that seven consonants, namely d, f, g, l, n, p and s, and five vowels a, a, e, o, 65, are used. These twelve phonetic sounds are easily distinguishable and permit the formation of a great number of different combinations, or sequences to which apparatus can be made to respond. Each of the above enumerated phonetic sounds is the sole representative of a group of similar sounds. There is no t which could not be distinguished from d if improperly pronounced, no v or b which could be taken for f or p, and for example, of the group of similar phonetic sounds appearing underlined in the following words, only the sound as in no is used: no, all, horn, hot.

Due to this selection, the representative selected phonetic sounds are more diiferent from each other than from the phonetic sounds of the group represented thereby. In accordance with the present invention each sequence is recognized and used for producing a command signal by which a particular operation of an apparatus is obtained.

An apparatus according to a preferred embodiment of the present invention comprises receiving means for receiving sequences of phonetic sounds and for converting the phonetic sounds into complex electrical waves; a plurality of selector means, each selector means producing an electrical command signal when receiving all the complex electrical waves to which the respective selector means is responsive, all selector means bing operatively connected to the receiving means; and a calculator for performing totalizing and other operations. The digital input means of the calculator are respectively connected to different selector means and actuated by command signals produced by the same, and the operating means of the calculator are operatively connected to other selector means and actuated by command signals produced by the same.

The speech controlled calculator apparatus of the present invention is preferably of the ten key type so that certain selected sequences of selected phonetic sounds will effect operations corresponding to the depression of a digital key, whereas other sequences of selected sounds will produce operations corresponding to the actuation of the totalizing key, the subtraction key, or of the nonaddition key.

Preferably, the apparatus is operated in accordance with the present invention so that, for example, the operating means by which totalizing and other operations of the machine are initiated, are actuated when selected short sequences of selected easily distinguishable phonetic sounds, not representing a word in an existent language, are spoken by the operator.

However, the apparatus of the present invention is not limited to calculating apparatus, and the present invention relates to any kind of apparatus which will respond only to selected sequences of selected phonetic sounds to produce command signals by which operations are controlled.

The novel features which are considered as characteristic for the invention are set forth in particular in the appended claims. The invention itself, however, both as to its construction and its method of operation, together with additional objects and advantages thereof, will be best understood from the following description of specific embodiments when read in connection with the accompanying drawings, in which:

FIG. 1 is a schematic circuit diagram illustrating the invention as embodied in a speech controlled computing apparatus; and

FIG. 2 is a schematic plan view of a calculator of the ten key type disclosed in the US. Patent 2,203,336, Landsiedel, published June 4, 1940, controlled by the circuit illustrated in FIG. 1.

Referring now to the drawing, sound receiving means, shown to be a microphone 50, are connected to a plurality of selector means 611 to 69, which are respectively associated with the digits from 0 to 9, and three other selector means 70, 71 and 72. Only the selector means 61 and 62 are shown in detail, and the other selector means are correspondingly constructed. Each selector means includes between two and four tuned means. Selector means 61 is shown to include three tuned

means

21, 22, 23, selector means 62 is shown to have two tuned

means

211, and 212 and selector means 63 are shown to have four tuned means 213 to 216.

The term tuned means as used in the present application defines a means which is tuned to a complex electrical wave corresponding to a phonetic sound spoken into the microphone 50. Such tuned means are disclosed in detail in the previously listed US. patents, and many consist of a number of band filter elements or include other refinements for assuring that only a single electric wave corresponding to a phonetic sound of the limited number of selected phonetic sounds passes through the respective tuned means 21. However, the tuned means may be so devised that it responds to several phonetic sounds of a group of similar phonetic sounds which are represented by a single phonetic sound. For example, the tuned means tuned to s may also respond to a lisped s (th), or to sh since these sounds are not used for operating other tuned means. A relay 31 is connected in series with tuned means 21 and constitutes a control element which controls a switch 311 connected in series with the tuned means 22, and a switch 312 connected into another circuit which includes a line connected to a source of electric current 90, and another line 91 connected to ground.

Another relay 32 is connected in series with tuned means 22 and switch 311, and controls switch 321 and switch 322. Switch 321 is connected in series with the third tuned means 23, and with a relay 33 controlling switch 332, which is connected in series with

switches

312 and 322. An electro-magnetic operating means 11, provided with a movable armature 111 is connected in series with the three

switches

312, 322, 332.

The selector means 62 is arranged in the same manner but only two tuned

means

211 and 212 are provided which are connected in series, respectively, with relays 34 and 35. Relay 34 controls switch means 341 and 342, and relay 35 controls switch means 352. An electro-magnetic operating means 12 including a movable armature 121 is connected in series with

switches

342 and 352. Selector means 63 is shown to include four tuned means 213 to 216, and the respective circuit of the selector means 63 includes relays and switches as described with reference to selector means 61, 62. An electro-magnetic operating means 13 is again provided. The selector means 64 to 69, 6t), 7t), 71 and 72 are not shown in detail, but are constructed corresponding to the above described selector means. Each selector means is associated with an electromagnetic operating means 14 to 19, 10, 80, 81 and 82, each provided with a movable armature.

FIG. 2 illustrates schematically the layout of a keyboard of a known computing machine of the ten key calculator type. For example, the computing machine schematically indicated in FIG. 2, may correspond to the calculator disclosed in the US. Patent 2,203,336, but any other computing machine may be used. The digit keys are replaced by the electromagnetic input means 10 to 19, or directly operated by the same, the operating means replaces or operates the totalizer key, the operating means 81 replaces or operates the subtraction key, and the operating means 82 replaces or operates the non-addition key. The arrangement may correspond to the actuator disclosed in the U.S. Patent 1,753,991. The apparatus operates in the following manner:

Referring first to the selector means 61, which is associated with the digit 1, the three tuned

means

21, 22 and 23 are respectively tuned to three different complex electrical waves which ar respectively associated with selected phonetic sounds. For example, using the phonetic system applied in Websters Dictionary, the three tuned

means

21, 22, 23 may be tuned to the complex electrical wave associated with the phonetic sounds, a, n, s, which together form the sequence ans.

Assuming that the phonetic sequence ans, which is not a word in an existent language, but which represents the digit 1, is spoken into the microphone 50, a complex electric wave corresponding to the phonetic sound a will pass through tuned means 21 and through relay 31. This will effect closing of

switches

311 and 312. The next spoken sound is n and a corresponding electric wave will pass through tuned means 22, switch 311, and relay 32 which closes switches 321 and 322.

If the sound n would have been spoken before the sound a, the corresponding electric wave could not have passed through tuned means 22, since switch 311 would not have been closed.

When the phonetic sound s is spoken, the corresponding electric wave passes through tuned means 23, which is possible since switch 321 is closed, and also passes through relay 33 which closes switch 332. Since all switches, 312, 322, and 332 which are connected in series with electro-rnagnetic operating means 11 are now closed, electro-magnetic operating means 11 is energized, and attracts its armature 111 whereby the digital operating means of the computing machine associated with the digit 1 is actuated. The selector means 62 is operated by a sequence of two phonetic sounds, for example, the sequence ddi. The electro-magnetic operating means 12 will only respond when the two phonetic sounds are spoken in the proper sequence, whereas the other selector means 62 to 69, 60 are constructed in the same manner and have at least two tuned means so that the respective digital input means of the computing machine are only actuated when a selected sequence of selected phonetic sounds is spoken into the microphone.

Particular sequences of phonetic sounds are associated with the totalizing, subtracting and non-addition opera tions of the calculator. For example, when the sequence is is spoken into the microphone, the selector means 70 will respond since it is provided with two corresponding tuned means, not shown, and effect operation of the electro-magnetic operating means 80 which performs in the computing machine an operation corresponding to the depression of the totalizing key in a computing machine.

Computing machines of this type are provided with printing means 104 and such printing means will print digits. It will be noted that in contrast to the arrangement provided in phonetic typewriters, not the individual sounds pronounced by the operators are printed, but a symbol representing an idea or information represented by a sequence of phonetic sounds.

If it is desired to add, starting with the highest order, two multi-order numbers having a different number of orders, for example, 321 and 21, it is only necessary to pronounce the sequence of phonetic sounds representing the digit before pronouncing the two sequences of phonetic sounds representing the digits 2 and 1. In this manner corresponding decimal orders will be properly associated in two computed numbers.

By way of example, it will now be shown how the digits from O to 9 may be expressed by sequences of easily distinguishable phonetic sounds.

1 an (ans) 2 d 3 s (send) 4 f6 (fon) 5 fa 6 g 7 66p 8 ad Total 1a Subtract s'zi Non-add go It will be noted that sequences of two phonetic sounds are adequate for representing the digits, and the totalizing, subtracting and non-addition operations of the computing machine. However, since in FIG. 1 the selector means 61, 63 and 64 are illustrated to have more than two tuning means, corresponding sequences of phonetic sounds are indicated in parenthesis in the table opposite digits 1, 3 and 4. It is evident that, although only thirteen easily distinguishable phonetic sounds are used, an extremely great number of combinations including between two and four of such phonetic sounds can be formed, and the number of such combinations and sequences is so great that for all practical purposes any operation in any machine can be controlled by spoken sequences of phonetic sounds in accordance with the present invention. Due to the fact that no existent language is used, the command words can be chosen very short so that the construction of the apparatus is simplified, and since only easily distinguishable phonetic sounds are used, an incorrect response of the apparatus due to non-recognition of the same letter pronounced by different speakers as distinct phonetic sounds, is made impossible.

The method and apparatus of the present invention have been described with reference to a calculator which is a particularly advantageous application of the present invention. However, it will be understood that any selected sequence of selected phonetic sounds can be used for actuating an electro-magnetic operating means for performing a function which is normally controlled by any manual operation.

A space traveler, confined in a capsule, and exposed to the forces of acceleration and deceleration, or becoming weightless, can operate a very great number of devices by spoken sequences of selected phonetic sounds, which do not correspond to an existent language, but are selected for the purposes of simplicity and accuracy.

It is not necessary that the sound receiving means is directly connected to the selector means, but the sound receiving means may control a transmitter for radio waves which are received at a distant point by the selector means. The element 51 in FIG. 1 may be considered to represent a transmitter and a receiver for radio waves.

In the schematic diagram on FIG. 1, holding circuits for the relays have been omitted for the sake of clarity. It will be understood by those skilled in the art, that the relays will hold the controlled switches in closed position, until the respective electro-magnetic operating means has performed its function whereupon the relay switches will return to their initial positions.

It will be understood, that the method and apparatus described above may also find a useful application for controlling other apparatus by the human voice. While the invention has been illustrated and described as embodied in a computing machine, it is not intended to be limited to the details shown, since structural modifications may be made without departing from the spirit of the present invention.

Without further analysis, the foregoing will so fully reveal the gist of the present invention that others can by applying current knowledge readily adapt it for various applications without omitting features that, from the standpoint of prior art, fairly constitute essential characteristics of the generic or specific aspects of this invention tion and, therefore, such adaptations should and are intended to be comprehended within the meaning and range of equivalence of the following claims.

I claim:

1. A speech controlled apparatus comprising, in com binati-on, a machine having a plurality of cooperating machine elements, and a plurality of operating means, each operating means being operable by a command signal to move from a normal inoperative condition to an actuated condition and being operatively connected with said plurality of machine elements, each operating means in said actuated condition causing said machine elements to cooperate in a different manner so that said machine performs different operations when different operating means are placed in said actuated condition thereof;

receiving means for receiving different sequences of at least two sound waves produced and selected by an operator to represent different operations of said machine, and for converting said sound waves into different sesuences of at least two electric waves respectively representing said different operations of said machine;

a plurality of selector means operatively connected with said receiving means for receiving said different sequences of at least two electric waves from the same, each selector means being responsive to only one selected sequence of said different sequences of electric Waves to produce a command signal so that the command signal of each selector means represents a different operation of the machine, and so that only one of said selector means produces a command signal representing one of said different operations when the operator selects said one of said ditferent operations by producing the respective sequence of sound waves,

each selector means being operatively connected with one of said operating means, respectively, for moving the same by said command signal thereof to said actuated condition whereby said machine is controlled by the operator to perform different operations.

2. A speech controlled apparatus comprising, in combination, a machine having a plurality of cooperating movable machine elements including a drive element for driving the other machine elements, and a plurality of operating means, each operating means being operable by a command signal to move from a normal inoperative position to an actuated position, and being operatively connected with said machine elements including said drive element, each operating means in said actuated position causing said machine elements to cooperate in a different manner While driven by said drive element so that said machine performs different operations when different operating means are placed in said actuated condition thereof;

receiving means for receiving different sequences of at least two phonetic sounds selected by an operator and constituting words respectively representing different operations of said machine, and for converting said different sequences of phonetic sounds into different sequences of at least two waves respectively representing said ditferent operations of said machine;

a plurality of selector means operatively connected with said receiving means for receiving said different sequences of at least two waves from the same, each select-or means being responsive to only one selected sequence of said different sequences of waves to produce a command signal so that the command signal of each selector means represents a different operation of the machine, and so that only one of said selector means produces a command signal representing one of said different operations when the operator selects said one of said different operations by sounding the sequence of phonetic sound which represents said one operation,

each selector means being operatively connected with one of said operating means, respectively, for moving the same by said command signal thereof to said actuated position whereby said machine is commanded by different words selected and spoken by the operator to said receiving means to perform different operations.

3. A speech controlled apparatus comprising receiving means for receiving different sequences of phonetic sounds respectively representing different operations, said receiving means converting said phonetic sounds into corresponding complex electrical Waves forming corresponding sequences of waves respectively representing said different operations;

a plurality of selector means, each selector means including a selected small number of tuned means tuned to different complex electrical waves having characteristic frequency components and corresponding to different and distinct phonetic sounds, and means responsive to a selected sequence of the respective complex electrical Waves to produce an electrical command signal when the respective selector means receives in a selected succession all the complex electrical waves to which the respective tuned means are responsive, said tuned means of each selector means being tuned to a different combination of said complex electrical waves, all said selector means being operatively connected to said receiving means to receive complex electrical waves from the same representing phonetic sounds so that any selector means produces a command signal only when said receiving means receives a sequence of phonetic sounds producing the sequence of complex electrical waves to which the respective selector means is responsive, spoken by an operator and selected by the same to represent selected operations;

a plurality of operating means respectively associated with said ditferent operations, each operating means being operatively connected with one of said selector means, respectively, to receive the respective command signal from the same, and being operable by the respective command signal to assume an actuated condition; and

apparatus controlled by each of said operating means in said actuated conditions to perform a different operation selected by the operator by sounding one of said selected sequences.

4. A speech controlled computing apparatus comprising receiving means for receiving different sequences of at least two phonetic sounds respectively representing different data and different computing operations, said receiving means converting said different sound sequences into different sequences of at least two complex electric waves respectively characteristic of said data and of said different computing operations;

a plurality of selector means, each selector means being responsive to only one of said different sequences of electric waves and including means for producing a command signal representing a selected data or computing operation only when receiving the sequence of complex electric waves to which the respective selector means is responsive,

said selector means being connected to said receiving means to receive said waves from the same when the same receives sequences of phonetic sounds spoken by an operator and selected by the same to represent selected data or computing operations;

a plurality of operating means, some of said operating means being associated with said data and other of said operating means being associated with said computing operations, each operating means being operatively connected with one of said selector means, respectively to receive a command signal from the same, and being operable by the respective command signal to assume an actuated condition;

computer means controlled by each of said operating means in said actuated condition to perform with data selected by the operator by sounding selected sequences of phonetic sounds, a ditferent computing operation selected by the operator by sounding another selected sequence; and

means for identify-ing the result data of the computing operations by perceptible representations of the result data.

5. Calculating apparatus controlled by human speech and comprising, in combination, receiving means for receiving sequences of phonetic sounds and for converting said phonetic sounds into corresponding complex electrical waves; a plurality of selector means, each selector means being responsive to at least two and to not more than a selected small number of different and distinct complex electrical waves having characteristic frequency components and corresponding to distinct phonetic sounds, and including means for producing an electrical command signal when receiving all the complex electrical waves to which the respective selector means is responsive, each selector means being responsive to a different combination of complex electrical waves, all said selector means being operatively connected to said receiving means to receive complex electrical waves representing phonetic sounds so that any selector means produces a command signal only when said receiving means receives a sequence of phonetic sound-s producing the plurality of complex electrical waves to which the respective selector means is responsive whereby said sequences of phonetic sounds need not represent words of an existent language; a calculator including ordinal calculating apparatus for per- =forming totalizing and other operations, digital input means controlling said calculating apparatus, and a plurality of operating means for actuating said calculating apparatus to perform with values introduced into said digital input means totalizing and other operations, said digital input means being respectively connected to different selector means of said selector means and actuated by command signals produce-d by the same, and said operating means being operatively connected to other selector means of said selector means and actuated by command signals produced by the same; and means operable by said calculating apparatus for identifying the results of said operations by perceptible representations of digits.

6. Calculating apparatus controlled by human speech and comprising, in combination, receiving means for receiving sequences of phonetic sounds and for converting said phonetic sounds into corresponding complex electrical waves; a plurality of selector means, each selector means being responsive to a different selected sequence of at least two and to not more than a selected small number of different and distinct complex electrical waves having characteristic frequency components and corresponding to distinct phonetic sounds, and including means for producing an electrical command signal when receiving all the complex electrical waves to which the respective selector means is responsive in the respective selected sequence, all said selector means being operatively connected to said receiving means to receive complex electrical waves representing phonetic sounds so that any selector means produces a command signal only when said receiving means receives a sequence of phonetic sounds producing the plurality of complex electrical waves to which the respective selector means is responsive whereby said sequences of phonetic sounds need not represent words of an existent language; a calculator including ordinal calculating apparatus for performing totalizing and other operations, ten digital input means controlling said calculating apparatus, and a plurality of operating means for actuating said calculating apparatus to perform with values introduced into said digital input means totalizing and other operations, said ten digital input means being respectively connected to ten of said select-or means and actuated by command signals produced by the same, and said operating means being operatively connected to other selector means of said selector means and actuated by command signals produced by the same; and means operable by said calculating apparatus for identifying the results of said operations by perceptible representations of digits.

7. A calculating apparatus controlled by human speech and comprising, in combination, receiving means for receiving sequences of phonetic sounds and for converting said phonetic sounds into corresponding complex electrical waves; a plurality of selector means, each selector means including a selected small number of tuned means tuned to different complex electrical waves having characteristic frequency components and corresponding to different and distinct phonetic sounds, and means responsive to a selected sequence of the respective complex electrical waves to produce an electrical command signal when the respective selector means receives in a selected succession all the complex electrical waves to which the respective tuned means are responsive, said tuned means of each selector means being tuned to a different combination of said complex electrical waves, all said selector means being operatively connected to said receiving means to receive complex electrical waves from the same representing phonetic sounds so that any selector means produces a command signal only when said receiving means receives a sequence of phonetic sounds producing the plurality of complex electrical Waves to which the respective selector means is responsive; a calculator including ordinal calculating apparatus for performing calculating operations, digital input means controlling said calculating apparatus, and operating means for actuating said calculating apparatus to perform with values introduced into said digital input means calculating other operations, said digital input means being respectively connected to different selector means and actuated by command signals produced by the same, and said operating means being operatively connected to other selector means of said selector means and actuated by command signals produced by the same; and means operable by said calculating apparatus for identifying the results of said operations by perceptible representations of digitsv 8. Calculating apparatus controlled by human speech and comprising, in combination, receiving means for receiving sequences of phonetic sounds and for converting said phonetic sounds into corresponding complex electrical waves; a plurality of selector means, each selector means including at least two and not more than a selected small number of tuned means tuned to different complex electrical waves having characteristic frequency components and corresponding to distinct phonetic sounds, and means for producing an electrical command signal when the respective selector means receives all the complex electrical waves to which the respective tuned. means are responsive, said tuned means of each selector means being tuned to a different combination of said complex electrical waves, all said selector means being operatively connected to said receiving means to receive complex electrical waves representing phonetic sounds so that any selector means produces a command signal only when said receiving means receives a sequence of phonetic sounds producing the plurality of complex electrical waves to which the respective selector means is responsive whereby said sequences of phonetic sounds need not represent Words of an existent language; a calculator including ordinal calculating apparatus for performing totalizing and other operations, ten digital input means controlling said calculating apparatus, and a plurality of operating means for actuating said calculating apparatus to perform with values introduced into said digital input means totalizing and other operations, said ten digital input means being respectively connected to ten of said selector means and actuated by command signals produced by the same, and said operating means being operatively connected to other selector means of said selector means and actuated by command signals produced by the same; and means operable by said calculating apparatus for identifying the results of said operations by perceptible representations of digits.

9. Calculating apparatus controlled by human speech and comprising, in combination, receiving means for receiving sequences of phonetic sounds and for converting said phonetic sounds into corresponding complex electrical waves; a plurality of selector means, each selector means including at least two and not more than a selected small number of tuned means tuned to different complex electrical waves having characteristic frequency components and corresponding to distinct phonetic sounds, and means responsive to a selected sequence of the respective electrical waves for producing an electrical command signal when the respective selector means receives all the com plex electrical waves to which the respective tuned means are responsive in the respective selected sequence, said tuned means of each selector means being tuned to a different sequence of said complex electrical waves, all said selector means being operatively connected to said receiving means to receive complex electrical waves representing phonetic sounds so that any selector means produces a command signal only when said receiving means receives a sequence of phonetic sound producing the plurality of complex electrical waves to which the respective selector means is responsive whereby said sequences of phonetic sounds need not represent words of an existent language; a calculator including ordinal calculating apparatus for performing totalizing and other operations, ten digital input means controlled said calculating apparatus, and a plurality of operating means for actuating said calculating apparatus to perform with values introduced into said digital input means totalizing and other operations, said ten digital input means being respectively connected to ten of said selector means and actuated by command signals produced by the same, and said operating means being operatively connected to other selector means of said selector means and actuated by command signals produced by the same; and printing means for identifying the results of said totalizing and other operations.

10. An apparatus for differentiating human speech and comprising receiving means for receiving sequences of phonetic sounds and for converting said phonetic sounds into corresponding complex electrical waves; a selector means operatively connected to said receiving means to receive complex electrical waves from same and including at least two and not more than a selected small number of tuned means tuned to different complex electrical waves having characteristic frequency components and corresponding to different and distinct phonetic sounds, at least two control elements, each control element being connected with one of said tuned means to be energized when a complex electrical wave passes through the same, a first switch and a second switch controlled by one of said control elements to be actuated when said one control element is energized, said first switch being connected with said other control element, and a third switch connected with said second switch and being controlled by the other control element to be actuated when the same is energized after actuation of said first switch by the energized first control element; a voltage source; and an electrically operated means connected to said second and third switches and to said voltage source so as to be energized when both said second and third switches are actuated whereby said electrically operated means is actuated only when said receiving means receives at least two selected phonetic sounds in a selected succession.

11. An apparatus for differentiating human speech and comprising receiving means for receiving sequences of phonetic sounds and for converting said phonetic sounds into corresponding complex electrical waves; a selector means operatively connected to said receiving means to receive complex electrical waves from the same and including at least two and not more than a selected small number of tuned means tuned to different complex electrical waves having characteristic frequency components and corresponding to different and distinct phonetic sounds, at least two relay windings, each relay winding being connected in series with one of said tuned means to be energized when a complex electrical wave passes through the same, .a first normally open switch and a second normally open switch controlled by one of said relay windings to close when said one relay winding is energized, said first switch being connected in series with said other relay Winding, and a third normally open switch connected in series with said second switch and being controlled by the other relay windin g to close when the same is energized after closing of said first switch by the energized first relay winding; a voltage source; and an electrically operated means connected in series to said second and third switches and to said voltage source so as to be energized when both said second and third switches are closed whereby said electrically operated means is actuated only when said receiving means receives at least two selected phonetic sounds in a selected succession.

12. An electrical apparatus operated by human speech and comprising receiving means for receiving sequences of phonetic sounds and for converting said phonetic sounds into corresponding complex electrical waves; a plurality of selector means operatively connected to said receiving means to receive complex electrical waves from the same, each selector means including at least two and not more than a selected small number of tuned means tuned to different complex electrical waves having characteristic frequency components and corresponding to different and distinct phonetic sounds, at lea-st two control elements, each control element being connected with one of said tuned means to be energized when a complex electrical wave passes through the same, a first switch and a second switch controlled by one of said control elements to be actuated when said one control element is energized, said first switch being connected in series with said other control element and, a third switch connected with said second switch and being controlled by the other control element to be actuated when the same is energized after actuation of said first switch by the energized first control element; a voltage source; and a plurality of electromagnetic operating means connected to said second and third switches of said selector means, respectively, and to said voltage source so as to be energized when both said second and third switches of the respective selector means are actuated whereby each of said electromagnetic means is actuated only when said receiving means receives at least two selected phonetic sounds in a selected succession.

13. A calculating apparatus controlled by human speech and comprising receiving means for receiving sequences of phonetic sounds and for converting said phonetic sounds into corresponding complex electrical waves; a plurality of selector means, each selector means including at least two and not more than a selected small number of tuned means tuned to different complex electrical waves having characteristic frequency components and corresponding to difierent and distinct phonetic sounds, at least two control elements, each control element being connected with one of said tuned means to be energized when a complex electrical wave passes through the same, a first switch and a second switch controlled by one of said control elements to be actuated when said one control element is energized, said first switch being connected in series with said other control element, and a third switch connected with said second switch and being controlled by the other control element to be actuated when the same is energized after actuation of said first switch by the energized first control element; a voltage source; .a plurality of electromagnetic operating means connected to said second and third switches of said selector means, respectively, and to said voltage source so as to be energized when both said second and third switches of the respective selector means are actuated whereby each of said electromagnetic means is actuated only when said receiving means receives at least two selected phonetic sounds in a selected succession; a calculator including ordinal calculating apparatus for performing totalizing and other operations, digital input means controlling said calculating apparatus, and a plurality of operating means for actuating said calculating apparatus to perform with values introduced into said digital input means totalizing and other operations, said digital input means and said operating means being, respectively, operatively connected to different electromagnetic means and actuated by the same; and means operable by said calculating apparatus for identifying the results of said operations by perceptible representations of digits. 7

14. A calculating apparatus controlled by human speech and comprising receiving means for receiving sequences of phonetic sounds and for converting said phonetic sounds into corresponding complex electrical waves; a plurality of selector means, each selector means including at least two and not more than a selected small number of tuned means tuned to difierent complex electrical waves having characteristic frequency components and corresponding to different and distinct phonetic sounds, at least two relay windings, each relay winding being connected in series with one of said tuned means to be energized when a complex electrical wave passes through the same, a first normally open switch and a second normally open switch controlled by one of said relay windings to close when said one relay Winding is energized, said first switch being connected in series with said other relay winding, and a third normally open switch connected in series with said second switch and being controlled by the other relay winding to close when the same is energized after closing of said first switch by the energized first relay winding; a voltage source; a plurality of electromagnetic operating means connected to said second and third switches of said selector means, respectively, and to said voltage source so as to be energized when both said second and third switches of the respective selector means are closed whereby each of said electromagnetic means is actuated only when said receiving means receives at least two selected phonetic sounds in a selected succession; a calculator including ordinal calculating apparatus for performing totalizing and other operations, digital input means controlling said calculating apparatus, and a plurality of operating means for actuating said calculating apparatus to perform with values introduced into said digital input totalizing and other operations, said digital input means and, said operating means being, respectively, operatively connected to difierent electromagnetic means and actuated by the same; and means operable by said calculating apparatus for identifying the results of said operations by perceptible representations of digits.

15. A method of performing calculating operations under the control of the human speech, comprising the steps of sounding selected sequences of at least two phonetic sounds so that different characteristic complex acoustic waves develop, ten of said sounded sequences representing digits and other sounded sequences representing commands for the performance of totalizing and other computing functions of a calculating apparatus; converting the acoustic wave of each sounded sequence of phonetic sounds into a command signal; and actuating by said command signals calculating apparatus to perform different operations under the control of ditferent command signals so that each sounded sequence causes either the introduction of a digit or the performance of a computing function.

16. A method of performing calculating operations under the control of the human speech, comprising the steps of sounding sequences of at least two phonetic sounds of a limited number of phonetic sounds, ten of said sounded sequences representing digits and other sounded sequences representing commands for the performance of totalizing and other computing functions of a calculating apparatus, said limited number of phonetic sounds being less than the number of phonetic sounds incident to a language, each phonetic sound of said limited number being representative of a group of similar phonetic sounds, the phonetic sounds of each said group being more similar to each other than to the phonetic sounds of the other groups so that said representative phonetic sounds of said limited number are phonetically more different from each other than from the phonetic sounds of the group represented thereby and have characteristic frequency components forming substantially different and distinct wave patterns; supplying said sounded sequences to apparatus responsive to said sounded sequences of said representative phonetic sounds and of similar phonetic sounds belonging, respectively, to the groups of phonetic sounds respectively represented by said representative sounds of each sounded sequence to produce a command signal; and actuating by said command signals calculating apparatus to perform different operations under the control of different command signals so that each sounded sequence causes either the introduction of a digit or the performance of a computing function.

References Cited by the Examiner UNITED STATES PATENTS 1,115,278 10/14 Bogstran-d 340-271 XR 1,547,226 7/25 Martin 340-171 1,949,136 2/34 Adair 179-1 2,203,336 6/40 Landsiedel 235-602 2,221,963 11/40 Case et al 104-151 2,575,909 11/51 Davis et a1 179-1 2,575,910 11/51 Mathes 179-1 2,613,273 /62 Kalfaian 179-1 2,646,465 7/53 Davis et al 179-16 2,685,615 8/54 Biddulph et al. 179-1 2,709,202 5/55- Handschin 179-6 2,921,133 1/60 Kalfaian 179-1 XR OTHER REFERENCES First Annual Summary Report on Speech Recognition, IBM Research Center, pp. 1-7, July 1, 1957.

MALCOLM A. MORRISON, Primary Examiner.

ABRAHAM BERLIN, WALTER W. BURNS, 11s.,

Examiners.

Claims

16. A METHOD OF PERFORMING CALCULATING OPERATIONS UNDER THE CONTROL OF THE HUMAN SPEECH, COMPRISING THE STEPS OF SOUNDING SEQUENCES OF AT LEAST TWO PHONETIC SOUNDS OF A LIMITED NUMBER OF PHONETIC SOUNDS, TEN OF SAID SOUND SEQUENCE REPRESENTING DIGITS AND OTHER SOUNDED SEQUENCES REPRESENTING COMMANDS FOR THE PERFORMANCE OF TOTALIZING AND OTHER COUPUTING FUNCTIONS OF A CALCULATING APPARATUS, SAID LIMITED NUMIBER OF PHONETIC SOUNDS BEING LESS THAN THE NUMBER OF PHONETIC SOUNDS INCIDENT TO A LANGUAGE, EACH OHONETIC SOUND OF SAID LIMITED NUMBER BEING REPRESENTATIVE OF A GROUP OF SIMILAR PHONETIC SOUNDS, THE PHONETIC SOUNDS OF EACH SAID GROUP BEING MORE SIMILAR TO EACH OTHER THAN TO THE PHONETIC SOUNDS OF THE OTHER GROUPS SO THAT SAID REPRESENTATIVE PHONETIC SOUNDS OF SAID LIMITED NUMBER ARE PHONETICALLY MORE DIFFERENT FROM EACH OTHER THAN FROM THE PHONETIC SOUNDS OF THE GROUP REPRESENTED THEREBY SAND HAVE CHARAC-