Methods and apparatus for silent speech interface
A system may detect silent, internal articulation of words by a human user, by measuring low-voltage electrical signals at electrodes positioned on a user's skin. The measured signals may have been generated by neural activation of speech articulator muscles during the internal articulation. The system may detect the content of internally articulated words even though the internal articulation may be silent, may occur even when the user is not exhaling, and may occur without muscle movement that is detectable by another person. The system may react in real-time to this detected content. In some cases, the system reacts by providing audio feedback to the user via an earphone or a bone conduction transducer. In other cases, the system reacts by controlling another device, such as a luminaire or television. In other cases, the system reacts by sending a message to a device associated with another person.
1. A method comprising:
(a) taking measurements of a set of electrical signals at positions on a user's skin, which skin is part of the user's head or neck, and which electrical signals are generated by neural activation of speech articulator muscles; and
(b) analyzing the measurements to recognize content of internally articulated speech by the user;
wherein at least a portion of the internally articulated speech occurs when the user is not exhaling.
2. The method of claim 1 , wherein the analyzing the measurements includes identifying temporal windows during which the electrical signals are low-voltage.
3. The method of claim 1 , wherein the analyzing the measurements includes identifying temporal windows during which each electrical signal, in the set of electrical signals, occurs at a specific position on the user's skin and has a root mean square (RMS) voltage, which RMS voltage: (a) is greater than or equal to 8 microvolts and less than or equal to 20 microvolts; and (b) is the RMS potential difference between voltage at the specific position and voltage at a reference electrode that is positioned on skin of an ear of the user.
4. The method of claim 1 , wherein the content which is recognized comprises one or more words.
5. The method of claim 1 , wherein the method further comprises providing audio feedback to the user, via sound vibrations produced by an earphone or bone conduction transducer.
6. The method of claim 5 , wherein the audio feedback comprises words that are at least part of the content of the internally articulated speech.
7. The method of claim 5 , wherein the audio feedback comprises words that are not part of the content of the internally articulated speech.
8. The method of claim 5 , wherein the audio feedback provides information requested by the internally articulated speech.
9. The method of claim 5 , wherein:
(a) the audio feedback provides information requested by the internally articulated speech; and
(b) the method further comprises obtaining the information from a computer server via the Internet.
10. The method of claim 1 , wherein the method further comprises controlling at least one device in accordance with instructions, which instructions were at least part of the content of the internally articulated speech.
11. The method of claim 1 , wherein the method further comprises sending a message that includes at least a portion of the content of the internally articulated speech.
12. The method of claim 1 , wherein the method further comprises:
(a) sending, to a device associated with a person other than the user, a first message that includes at least a portion of the content of the internally articulated speech;
(b) receiving a second message from the device;
(c) recognizing content of the second message; and
(d) providing audio feedback to the user, which audio feedback comprises words that are part of the content of the second message.
13. The method of claim 1 , wherein the measurements comprise measurements of electrical voltage with electrodes.
14. The method of claim 13 , wherein:
(a) the electrodes comprise one or more reference electrodes and one or more other electrodes; and
(b) the other electrodes are positioned on skin of the user in one or more of the following skin regions: (i) mental; (ii) oral; (iii) infraorbital; (iv) buccal; (v) submental; (vi) submaxillary;
(vii) hyoid; and (viii) carotid fossa.
15. The method of claim 14 , wherein the one or more reference electrodes are positioned on regions of skin of the user, which regions are part of one or more ears of the user or of one or more arms of the user.
16. The method of claim 1 , wherein the analyzing the measurements includes inputting data into one or more neural networks, where the data encodes the measurements.
17. Apparatus comprising:
(a) electrodes; and
(b) one or more computers;
wherein
(i) the electrodes are configured to take, at positions on a user's skin, measurements of a set of electrical signals that are generated by neural activation of speech articulator muscles, and
(ii) the one or more computers are programmed to analyze the measurements to recognize content of internally articulated speech by the user.
18. The apparatus of claim 17 , wherein the one or more computers are programmed to identify temporal windows during which the electrical signals are low-voltage.
19. The apparatus of claim 17 , wherein the one or more computers are programmed to identify temporal windows during which each electrical signal, in the set of electrical signals, occurs at a specific position on the user's skin and has a root mean square (RMS) voltage, which RMS voltage: (a) is greater than or equal to 8 microvolts and less than or equal to 20 microvolts; and (b) is the RMS potential difference between voltage at the specific position and voltage at a reference electrode that is one of the electrodes and is positioned on skin of an ear of the user.
20. The apparatus of claim 17 , wherein:
(a) the apparatus further comprises a transducer, which transducer comprises an earphone or bone conduction transducer; and
(b) the transducer is configured to provide audio feedback to the user.