DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application is being examined under the pre-AIA first to invent provisions.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
Claim(s) 1, 3-4, 6-9, 11, 13-18, 20 and 22-27 is/are rejected under 35 U.S.C. 103 as being unpatentable over Wolf et al. (US 2021/0142782) in view of Pereira et al. (US 10,922,484).
Claim 1,
Wolf teaches a method, comprising: receiving a set of text ([0064] referring to 216, and in some embodiments, a text input can be received; the device 102 of a first user can receive a text (e.g. text input, text message);
receiving initial audio content substantially corresponding to a voice of a user, associated with the set of text and generated by a text-to-speech (TTS) model that was trained using training data that includes audio of the user ([0003] [0060] [0067-0068] the voice model can be generated and further trained using audio samples provided by respective user; the voice model 140 can correspond to a text to speech (TTS) digital clone; the device 102 can convert the input text to a synthetic audio output 142 to model a voice of the second user; the device 102 can recite each word, phrase or sentence included in the received text using the synthetic audio output 142 corresponding to or resembling the second user’s voice and generated by the voice model 140).
The difference between the prior art and the claimed invention is that Wolf does not explicitly teach identifying, based on analysis of the initial audio content, a subset of text from the set of text that is to be recorded by the user; causing the subset of text to be recorded by the user to generate a representation of a user recording; and causing portions of the initial audio content associated with the subset of text to be updated by replacing the portions of the initial audio content with the user recording to define updated audio content that includes the user recording.
Pereira teaches identifying, based on analysis of the initial audio content, a subset of text from the set of text that is to be recorded by the user ([col. 2 lines 12-16] [col. 21 lines 31-32] when the system receives audio data, the system performs speech-to-text transcription on the audio data; the words of the transcribed audio are compared to the words of the comparison document, and the differences in words is flagged as errors; at 824, the comparator module 118 may select a portion of the script to re-record);
causing the subset of text to be recorded by the user to generate a representation of a user recording ([col. 12 lines 23-25 and lines 45-46] the narrator module 220 may prompt the narrator to re-record that particular section; at 826, the narrator module 220 may receive revised audio data of the portion of the script); and
causing portions of the initial audio content associated with the subset of text to be updated by replacing the portions of the initial audio content with the user recording to define updated audio content that includes the user recording ([col. 10 lines 47-58] [col. 21 line 59 to col. 22 line 9] the comparator module 118 may replace a particular defective section of a prior recording and receive a revised recording to be inserted; the narrator module 220 may insert revised audio data into the audio data; the revised recording may be inserted into the particular section of the recording).
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the teachings of Wolf with teachings of Pereira by modifying the generating a voice model for a user as taught by Wolf to include identifying, based on analysis of the initial audio content, a subset of text from the set of text that is to be recorded by the user; causing the subset of text to be recorded by the user to generate a representation of a user recording; and causing portions of the initial audio content associated with the subset of text to be updated by replacing the portions of the initial audio content with the user recording to define updated audio content that includes the user recording as taught by Pereira for the benefit of determining to re-re-record specific sections of the manuscript (Pereira [Abstract]).
Claim 3,
Wolf further teaches the method of claim 1, wherein: the receiving the set of text includes receiving the set of text at a processor of a first compute device ([0028] [0064] the device 102 can include one or more processors 104; the device 102 of a first user can receive a text (e.g. text input, text message)).
Pereira further teaches the identifying the subset of text includes: sending, via the processor of the first compute device, the initial audio content to a second compute device different from the first compute device ([col. 4 lines 23-39] [col. 4 lines 59 to col. 5 line 17] [col. 12 lines 17-41] the service provider 108 may include one or more content server(s) 110; the narrator UI may allow the user(s) 102 to provide the audio data; the narrator module 220 may transmit the recording to the content server(s) 110 for the proofing process) to cause the second compute device to generate an indication of the subset of text ([col. 21 lines 31-44] the comparator module 118 may select a portion of the script to re-record) based on at least one of human input or a software model stored at the second compute device ([col. 6 line 64 to col 7 line 34] [col. 9 lines 38-58] [col. 12 line 53 to col. 13 line 2] the computing architecture 200 of the content server(s) 110; computer-readable media 204 may store the data store 222; machine learning models stored in the data store 222; the auditor module 120 may present to the proofer to verify the defect); and
receiving, at the processor of the first compute device and from the second compute device, the indication of the subset of text ([col. 4 line 59 to col. 5 line 30] the narrator UI may display instructions to re-record portions of a narrating assignment due to defects found in the audio data; the user(s) 102 may interact with the service provider 108 to receive the tagged transcription; the auditor UI may display portions of the manuscripts and corresponding transcribed audio data with errors tagged).
Claim 4,
Wolf further teaches the method of claim 1, wherein the receiving the set of text includes receiving the set of text at a processor of a first compute device ([0028] [0064] the device 102 can include one or more processors 104; the device 102 of a first user can receive a text (e.g. text input, text message)), the method further comprising: to cause the TTS model to be retrained based on the updated audio content ([0005] [0070-0071] the method can include training the voice model for the user using one or more subsequent audio samples from the user; the machine learning model can refine the voice model 140; the decoder 130 can train and/or refine the voice model 140 for the user).
Pereira further teaches sending, from the processor of the first compute device and to a second compute device, a signal having the updated audio content ([col. 4 lines 23-39] [col. 4 lines 59 to col. 5 line 17] [col. 12 lines 17-41] [col. 17 lines 27-53] the service provider 108 may include one or more content server(s) 110; the narrator UI may allow the user(s) 102 to provide the audio data; the narrator module 220 may transmit the recording to the content server(s) 110; the revised recording may be inserted into the particular section of the recording).
Claim 6,
Pereira further teaches the method of claim 1, further comprising: receiving, at a recording identification Al model, a third audio content ([col. 9 line 19-58] the speech conversion module 214 may use the system model and audio data; a machine learning model, once trained, is a learned mechanism that can receive new data as input; the unknown audio may include audio recordings); and
outputting, from the recording identification Al model, an indication of a portion of the third audio content to be updated by the user ([col. 2 lines 8-25] [col. 4 line 59 to col. 5 line 17] [col. 21 lines 31-44] the difference in words is flagged as errors; the narrator UI may display instructions to re-record portions of a narrating assignment due to defects found in the audio data; the comparator module 118 may select a portion of the script to re-record).
Wolf teaches from the TTS model and/or a preprocessing Al model (0005] [0060] [0067] the voice model 140 can correspond to a text to speech (TTS) digital clone; generate, using the voice model a synthetic audio output; convert the input text to a synthetic audio output 142).
Claim 7,
Wolf teaches an apparatus, comprising: a processor; and a memory coupled to the processor, the memory storing instructions that when executed cause the processor to ([0027-0028] [0040-0041] device 102; processors 104; storage device 106):
receive a set of text ([0064] a text input can be received; the device 102 of a first user can receive a text (e.g. text input, text message));
generated by a text-to-speech (TTS) model that was trained using training data that includes audio of a user and that generates output substantially corresponding to a voice of the user ([0003] [0060] [0067-0068] the voice model can be generated and further trained using audio samples provided by the respective user; the voice model 140 can correspond to a text to speech (TTS) digital clone; convert the input text to synthetic audio output 142 to model a voice of the second user; the device 102 can recite each word phrase or sentence included in the received text using the synthetic audio output 142 corresponding to or resembling the second user’s voice);
The difference between the prior art and the claimed invention is that Wolf does not explicitly teach receive initial audio content associated with the set of text; cause, based on analysis of the initial audio content, a subset of text from the set of text to be recorded by the user to generate a representation of a user recording; and cause portions of the initial audio content associated with the subset of text to be updated by replacing the portions of the initial audio content with the user recording to define updated audio content that includes the user recording.
Pereira teaches receive initial audio content associated with the set of text ([col. 14 lines 1-19] the transceiver module 116 may receive audio data corresponding to the script);
cause, based on analysis of the initial audio content, a subset of text from the set of text to be recorded by the user to generate a representation of a user recording ([col. 2 lines 8-25] [col. 12 lines 16-41] [col. 21 lines 31-58] the system performs speech to text transcription on the audio data; the words of the transcribed audio are compared to the words of the comparison document, and the differences in words is flagged as errors; the comparator module 118 may select a portion of the script to re-record; the narrator module 220 may prompt the narrator to re-record that particular section; the narrator module 220 may receive revised audio data of the portion of the script); and
cause portions of the initial audio content associated with the subset of text to be updated by replacing the portions of the initial audio content with the user recording to define updated audio content that includes the user recording ([col. 10 lines 47-58] [col. 21 line 45 to col. 22 line 8] the comparator module 118 may replace a particular defective section of a prior recording and receive a revised recording to be inserted; the narrator module 220 may insert revised audio data into the audio data; the revised recording may be inserted into the particular section of the recording).
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the teachings of Wolf with teachings of Pereira by modifying the generating a voice model for a user as taught by Wolf to include initial audio content associated with the set of text; cause, based on analysis of the initial audio content, a subset of text from the set of text to be recorded by the user to generate a representation of a user recording; and cause portions of the initial audio content associated with the subset of text to be updated by replacing the portions of the initial audio content with the user recording to define updated audio content that includes the user recording as taught by Pereira for the benefit of determining to re-re-record specific sections of the manuscript (Pereira [Abstract]).
Claim 8,
Pereira further teaches the apparatus of claim 7, wherein: the processor is associated with a first compute device ([Figs. 1-2] [col. 4 lines 23-39] [col. 7 lines 1-7] the service provider 108 may include one or more content servers 110; the computing architecture 200 may include one or more processors), and
the instructions that cause the processor to cause the subset of text from the set of text to be recorded by the user include instructions that cause the processor to (1) send, to a second compute device, a signal that causes, based on the analysis of the initial audio content, the subset of text to be recorded by the user at the second compute device to generate the representation of the user recording ([col. 4 line 40 to col. 5 line 17] [col. 12 lines 17-41] the service provider 108 described herein may cause one or more user interfaces to be presented to users 102 via devices 104; the narrator UI may display instructions associated with a current narrating assignment; the instructions may include the text of a portion of manuscript; the narrator UI may display instructions to re-record portions of a narrating assignment due to defects found in the audio data; the narrator module 220 may prompt the narrator to re-record that particular section) and
(2) receive, from the second compute device, the representation of the user recording ([col. 11 line 61 to col. 12 line 41] [col. 14 lines 1-19] the narrator may use the recording tool to record the manuscript then upload the audio data from the narrator UI; the narrator module 220 may transmit the recording to the content servers 110 for the proofing process; the transcriber module 116 may receive audio data corresponding to the script).
Claim 9,
Wolf further teaches the apparatus of claim 7, wherein: the processor is associated with a first compute device ([0028] device 102 can include one or more processors 104).
Pereira further teaches the instructions that cause the processor to cause the portions of the initial audio content to be updated include instructions that cause the processor to send, to a second compute device, the portions of the initial audio content and the user recording to cause the second compute device to replace the portions of the initial audio content with the user recording to define the updated audio content that includes the user recording ([col. 10 line 47 to col. 11 line 15] [col. 12 lines 17-41] [col. 16 lines 22-36] [col. 21 line 45 to col. 22 line 8] the processor 400 may be performed by the user device 104 in cooperation with any one or more of the content servers 110; the narrator may upload the audio data from the narrator UI; the narrator module 220 may transmit the recording to the content servers 110 for proofing process; the narrator module 220 may provide the audio data recorded for the section that proceeds the particular section; the narrator module 220 may receive the revised recording; receive revised audio data of the portion of the script; replace a particular defective section of a prior recording and receive a revised recording to the be inserted; the narrator module 220 may insert revised audio data into the audio data).
Claim 11,
The apparatus of claim 7, wherein: the processor is associated with a first compute device, and the memory storing further instructions that when executed cause the processor to send, to a second compute device and after causing the portions of the initial audio content to be updated to define the updated audio content, a signal having the updated audio content to cause the TTS model to be retrained at the second compute device and based on the updated audio content. (Claim 11 contains subject matter similar to claim 4, and thus is rejected under similar rationale)
Claim 13,
The apparatus of claim 7, wherein: the memory storing further instructions that when executed cause the processor to: receive, at a recording identification AI model, a third audio content from the TTS model and/or a preprocessing AI model; and output, from the recording identification AI model, an indication of a portion of the third audio content to be updated by the user. (Claim 13 contains subject matter similar to claim 6, and thus is rejected under similar rationale)
Claim 14,
Wolf further teaches the apparatus of claim 7, wherein: the TTS model is included within a plurality of TTS models, the user is included within a plurality of users, each TTS model from the plurality of TTS models is trained with audio of a user from the plurality of users and not from remaining users from the plurality of users, each TTS model from the plurality of TTS models is uniquely associated with a genre from a plurality of genres and is configured to generate output substantially corresponding to a voice from the plurality of users used to train that TTS model ([0056-0072] the model is stored with a plurality of user voice models trained with audio uniquely cloned from a corresponding target user unique from the other users; a uniquely associated generic to personalized voice quality classes resembling a respective voice from the user to refine TTS training).
Claim 15,
Wolf further teaches the apparatus of claim 7, wherein: the TTS model is included within a plurality of TTS models, the user is included within a plurality of users, each TTS model from the plurality of TTS models is trained with audio of a user from the plurality of users and not any other user, each TTS model from the plurality of TTS models is configured to generate output substantially corresponding to a voice from the plurality of users used to train that TTS model ([0056-0072] the model is stored with a plurality of user voice models trained with audio uniquely trained from a corresponding target user unique from other users; each voice model outputs audio resembling a respective voice from the user that to refined TTS training).
Claim 16,
A processor-readable medium storing instructions that, when executed by a processor, cause the processor to: receive a set of text; receive initial audio content associated with the set of text and generated by a text-to-speech (TTS) model that was trained using training data that includes audio of a user and that generates output substantially corresponding to a voice of the user; cause, based on analysis of the initial audio content, a subset of text from the set of text to be recorded by the user to generate a representation of a user recording; and cause portions of the initial audio content associated with the subset of text to be updated by replacing the portions of the initial audio content with the user recording to define updated audio content that includes the user recording. (Claim 16 contains subject matter similar to claim 7, and thus is rejected under similar rationale)
Claim 17,
The processor-readable medium of claim 16, wherein: the processor is associated with a first compute device and the instructions that cause the processor to cause the subset of text from the set of text to be recorded by the user include instructions that cause the processor to (1) send, to a second compute device, a signal that causes, based on the analysis of the initial audio content, the subset of text to be recorded by the user at the second compute device to generate the representation of the user recording and (2) receive, from the second compute device, the representation of the user recording. (Claim 17 contains subject matter similar to claim 8, and thus is rejected under similar rationale)
Claim 18,
The processor-readable medium of claim 16, wherein: the processor is associated with a first compute device, and the instructions that cause the processor to cause the portions of the initial audio content to be updated include instructions that cause the processor to send, to a second compute device, the portions of the initial audio content and the user recording to cause the second compute device to replace the portions of the initial audio content with the user recording to define the updated audio content that includes the user recording. (Claim 18 contains subject matter similar to claim 9, and thus is rejected under similar rationale)
Claim 20,
The processor-readable medium of claim 16, wherein: the processor is associated with a first compute device, and the instructions further includes instructions that when executed cause the processor to send, to a second compute device and after causing the portions of the initial audio content to be updated to define the updated audio content, a signal having the updated audio content to cause the TTS model to be retrained at the second compute device and based on the updated audio content. (Claim 20 contains subject matter similar to claim 4, and thus is rejected under similar rationale)
Claim 22,
The processor-readable medium of claim 16, wherein: the instructions further including instructions that when executed cause the processor to: receive, at a recording identification AI model, a third audio content from the TTS model and/or a preprocessing AI model; and output, from the recording identification AI model, an indication of a portion of the third audio content to be updated by the user. (Claim 22 contains subject matter similar to claim 6, and thus is rejected under similar rationale)
Claim 23,
The processor-readable medium of claim 16, wherein: the TTS model is included within a plurality of TTS models, the user is included within a plurality of users, each TTS model from the plurality of TTS models is trained with audio of a user from the plurality of users and not any other user, each TTS model from the plurality of TTS models is configured to generate output substantially corresponding to a voice from the plurality of users used to train that TTS model. (Claim 23 contains subject matter similar to claim 15, and thus is rejected under similar rationale)
Claim 24,
Pereira further teaches the apparatus of claim 7, wherein the subset of text is associated with a portion of the initial audio content having at least one quality that is identified as less than acceptable, based on the analysis of the initial audio content ([col. 11 line 15 to col. 12 line 41] [col. 14 line 42 to col. 15 line 33] the comparator module 118 may flag common errors (e.g. consistency, pronunciation, spacing issue, noise, wrong word etc.); based on the analysis, if the common errors module 218 determines that the level of defects in the audio quality (e.g. excessive background noise, low or no volume, distortion in audio etc.) is below a threshold audio quality; if the proofing process determines that a particular section of the recording has a level of defects that meets the severity threshold, the narrator module 220 may prompt the narrator to re-record that particular section).
Claim 25,
Pereira further teaches the apparatus of claim 7, wherein the instructions that cause the processor to cause the subset of text from the set of text to be recorded by the user include instructions that cause the processor to: receive, from a compute device, a signal that is indicative of a portion of the initial audio content to be edited ([col. 4 lines 12-23] [col. 5 lines 18-30] [col. 20 lines 50-60] [claim 9] the users 102 may utilize devices 104 to interact with a service provider 108; the auditor UI may display portions of the manuscripts and corresponding transcribed audio data with errors tagged; the auditor module 120 may also present an audio tool to play the audio associated with the defective portion for verification; receiving user input to accept the one or more defects), and
retrieve the subset of text from the set of text, the subset of text being associated with the portion of the initial audio content ([col. 20 line 61 to col. 21 line 4] [claims 9 and 10] identifying a portion of the manuscript corresponding to a location of the one or more defects; presenting, in the user interface, the portion of the manuscript corresponding to the location of the one or more defects; the auditor module 120 may present the original manuscript and the tagged transcribed audio next to each other).
Claim 26,
Pereira further teaches the apparatus of claim 7, wherein the instructions that cause the processor to cause the subset of text from the set of text to be recorded by the user include instructions that cause the processor to: provide the initial audio content as input to a machine learning model to cause the machine learning model to identify a portion of the initial audio content to be edited ([claims 11 and 13] [col. 9 lines 19-58] the speech conversion module 214 may use the system model and audio data; a machine learning model, once trained, is a learned mechanism that can receive new data as input; receiving audio data corresponding a portion of the manuscript; identifying one or more differences as one or more defects; analyzing the audio data for an additional defect for an additional defect and determining to re-record the portion of the manuscript), and
retrieve the subset of text from the set of text, the subset of text being represented by the portion of the initial audio content ([claims 9-10 and 15] identifying a portion of the manuscript corresponding to a location of the one or more defects; presenting the portion of the manuscript corresponding to the location of the one or more defects; presenting a prompt to re-record the portion of the manuscript).
Claim 27,
Pereira further teaches the apparatus of claim 7, wherein the subset of text is associated with a portion of the initial audio content having a quality lower than a quality of at least one remaining portion of the initial audio content, based on the analysis of the initial audio content, the quality of the portion of the initial audio content being indicated by at least one of audio smoothness, audio clarity or audio tone ([col. 11 line 15 to col. 16 line 41] [col. 17 lines 27-53] [col. 21 lines 31-58] if the proofing process determines that a particular section of the recording has a level of defects that meets the severity threshold, the narrator module 220 may prompt the narrator to re-record that particular section; the narrator module 220 may provide the audio data recorded for the section that proceeds the particular section to match the volume and tone; determine that the tone for the revised recording matched the tone in the audio portions before and after the particular section; the level of defects in the audio quality (e.g. excessive background noise, low or no volume, distortion in audio etc.)).
Claim(s) 2 is/are rejected under 35 U.S.C. 103 as being unpatentable over Wolf et al. (US 2021/0142782) in view of Pereira et al. (US 10,922,484) and further in view of Lester (US 8,872,265).
Claim 2,
Wolf teaches the method of claim 1, wherein: the receiving the set of text includes receiving the set of text at a processor of a first compute device ([0028] [0064] the device 102 can include one or more processors 104; the device 102 of a first user can receive a text (e.g. text input, text message)).
The difference between the prior art and the claimed invention is that Wolf nor Pereira explicitly teaches the receiving the initial audio content includes receiving the initial audio content from a second compute device that is different from the first compute device and that stores the TTS model.
Lester teaches the receiving the initial audio content includes receiving the initial audio content from a second compute device that is different from the first compute device and that stores the TTS model ([col. 3 line 49 to col. 4 line 3] [col. 4 line 61 to col. 5 line 9] [col. 6 line 65 to col. 7 line 28] the user computing device 102 may transmit a desired customization over the network 106 to the content customization server 110 and may also receive modified or customized content over network 106 from the content customization server 110; additional operations of the content customization server 110, such as performing TTS conversions; the memory 210 may include a content customization module 216 to synthesize speech).
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the teachings of Wolf with teachings of Lester by modifying the generating a voice model for a user as taught by Wolf to include the receiving the initial audio content includes receiving the initial audio content from a second compute device that is different from the first compute device and that stores the TTS model as taught by Lester for the benefit of combining audio portions to produce a combined item of audio content with multiple voices (Lester [Abstract]).
Claim(s) 5, 12 and 21 is/are rejected under 35 U.S.C. 103 as being unpatentable over Wolf et al. (US 2021/0142782) in view of Pereira et al. (US 10,922,484) and further in view of Arik et al. (US 11,238,843).
Claim 5,
Wolf and Pereira teach all the limitations in claim 1. The difference between the prior art and the claimed invention is that Wolf nor Pereira explicitly teach wherein the set of text is a second set of text, the method further comprising: receiving, at a preprocessing artificial intelligence (Al) model, a first set of text; revising, at the preprocessing Al model, the first set of text to generate the second set of text to improve a quality of the initial audio content generated by the TTS model; and outputting, from the preprocessing Al model, the second set of text.
Arik teaches wherein the set of text is a second set of text ([col. 25 line 22 to col. 26 line 12] text preprocessing can be important for good performance; these issues may be alleviated by normalizing the input text; every word is replaced with it phoneme representation), the method further comprising:
receiving, at a preprocessing artificial intelligence (Al) model, a first set of text (col. 25 line 23 to col. 26 line 15] feeding raw text (characters with spacing and punctuation); a phoneme-only model requires a preprocessing step to convert words to their phoneme representations by using a separately trained grapheme-to-phoneme model; the text embedding model 2610 may comprise a phoneme-only model and/or mixed character-and-phoneme model);
revising, at the preprocessing Al model, the first set of text to generate the second set of text to improve a quality of the initial audio content generated by the TTS model ([col. 25 line 23 to col. 26 line 12] a phoneme-only model requires a preprocessing step to convert words to their phoneme representations; every word is replaced with its phoneme representation; it was found that this improves pronunciation accuracy and minimizes attention errors; text preprocessing can be important for good performance); and
outputting, from the preprocessing Al model, the second set of text ([col. 25 line 47 to col. 26 line 15] a phoneme-only model requires a preprocessing step to convert words to their phoneme representations; these models may be identical to character-only models except that the input layer of the encoder sometimes receives phoneme and phoneme stress embeddings instead of character embeddings; every word is replaced with its phoneme representation).
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the teachings of Wolf with teachings of Lester by modifying the generating a voice model for a user as taught by Wolf to include wherein the set of text is a second set of text, the method further comprising: receiving, at a preprocessing artificial intelligence (Al) model, a first set of text; revising, at the preprocessing Al model, the first set of text to generate the second set of text to improve a quality of the initial audio content generated by the TTS model; and outputting, from the preprocessing Al model, the second set of text as taught by Arik for the benefit of achieving good performance in terms of naturalness of the speech and its similarity to original speaker — even with very few cloning audios (Arik [Abstract]).
Claim 12,
The apparatus of claim 7, wherein: the set of text is a second set of text, the memory storing further instructions that when executed cause the processor to: receive, at a preprocessing artificial intelligence (AI) model, a first set of text; revising, at the preprocessing AI model, the first set of text to generate the second set of text to improve a quality of the initial audio content generated by the TTS model; and output, from the preprocessing AI model, the second set of text. (Claim 12 contains subject matter similar to claim 5, and thus is rejected under similar rationale)
Claim 21,
The processor-readable medium of claim 16, wherein: the set of text is a second set of text, the instructions further including instructions that when executed cause the processor to: receive, at a preprocessing artificial intelligence (AI) model, a first set of text; revise, at the preprocessing AI model, the first set of text to generate the second set of text to improve a quality of the initial audio content generated by the TTS model; and output, from the preprocessing AI model, the second set of text. (Claim 21 contains subject matter similar to claim 5, and thus is rejected under similar rationale)
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Kibre et al. (US 2002/0193994) – A new speaker provides speech from which comparison snippets are extracted. The comparison snippets are compared with initial snippets stored in a recorded snippet database that is associated with a concatenative synthesizer. The comparison of the snippets to the initial snippets produces required sound units. A greedy selection algorithm is performed with the required sound units for identifying the smallest subset of the input text that contains all of the text for the new speaker to read. The new speaker then reads the optimally selected text and sound units are extracted from the human speech such that the recorded snippet database is modified and the speech synthesized adopts the voice quality and characteristics of the new speaker.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHREYANS A PATEL whose telephone number is (571)270-0689. The examiner can normally be reached Monday-Friday 8am-5pm PST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Pierre Desir can be reached at 571-272-7799. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
SHREYANS A. PATEL
Primary Examiner
Art Unit 2653
/SHREYANS A PATEL/ Examiner, Art Unit 2659
Prosecution Insights