DETAILED ACTION
This office action is a First Action on the Merits (FAOM) for the claim set entered on 12/20/2024. Claims 1-7 are pending and have been considered.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA
Priority
Acknowledgment is made of applicant’s claim for foreign priority under 35 U.S.C. 119
(a)-(d). The certified copy has been filed for the parent Application No. JP2022-029327, filed on 02/28/2022.
Information Disclosure Statement
The information disclosure statement(s) submitted on 07/23/2024 is/are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement(s) is/are being considered by the examiner.
Drawings
Figure 1 should be designated by a legend such as --Prior Art-- because only that which is old is illustrated. See MPEP § 608.02(g). Corrected drawings in compliance with 37 CFR 1.121(d) are required in reply to the Office action to avoid abandonment of the application. The replacement sheet(s) should be labeled “Replacement Sheet” in the page header (as per 37 CFR 1.84(c)) so as not to obstruct any portion of the drawing figures. If the changes are not accepted by the examiner, the applicant will be notified and informed of any required corrective action in the next Office action. The objection to the drawings will not be held in abeyance.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
Such claim limitation(s) is/are: “a converting means for converting natural language text” and “a training means for training a language model” in claim 1, “training data forming means for forming training data” and “a pre-training means for pre-training said language model” in claim 2, “a noise adding means for adding noise”, “a training data forming means for forming training data for fine-tuning”, and “a fine-tuning means for fine-tuning” in claim 3, “a noise adding means for adding noise”, “a training data forming means for forming training data”, and “a fine-tuning means for fine-tuning said pre-trained language model” in claim 4, and “a noise adding means for adding noise”, “an additional training data forming means for forming additional training data”, and “an additional pre-training means for additionally pre-training said pre-trained language model” in claim 6.
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof. Specifically, the specification provides no additional definition as to what “means” is reasonably understood to represent in this context. The specification appears to define similar components to those claimed as “units”, wherein Fig. 6 shows an example “unit” for the noise-adding unit. The unit appears to be comprised of other “units” and/or components which are similarly lacking definition and/or physical structure. Fig. 16 defines a “computer system realizing the language model training device in accordance with the present invention”, wherein Fig. 17 shows the “hardware configuration of the computer system shown in Fig. 16”. The components of Fig. 17 will be understood to be capable of performing the actions of all means and/or units.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claim(s) 1-7 is/are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Independent claim 1 recites:
A language model training device, comprising:
a converting means for converting natural language text to output a sequence of phonetic letters; and
a training means for training a language model using said text and said sequence of phonetic letters output from said converting means.
Independent claim 6 recites:
A dialogue device realizing speech-based dialogue with a user, comprising:
a trained language model generated by machine learning using at least natural language text and a sequence of phonetic letters obtained by converting the text;
a semantic interpretation module with said trained language model, for receiving as an input speech information of said user; and
an utterance/response module for receiving as an input the speech information of said user and for executing a dialogue with the user under control of said semantic interpretation module.
Independent claim 7 recites:
A trained language model generated by machine learning, using at least natural language text and a sequence of phonetic letters obtained by converting the text.
These limitations, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. For example, the claim(s) read(s) on converting natural language text to phonemes, training a language model based on the text to phoneme conversion, and receiving speech information from a user to be responded to (claim 6). That is, other than reciting “language model”, nothing in the claim element precludes the step from practically being performed in the mind.
For example, regarding the conversion of text to phonemes, the examiner respectfully asserts that this is a mental process associated with reading/speaking. As a user is provided with a text, they will mentally determine how to pronounce words based on format, spelling, etc. These pronunciations can be written down in conjunction with the text. “Training” a language model to perform this conversion is applying a generic computing component to perform a mental process. Further, with regard to the specific situation of a conversation of claim 6, the examiner respectfully asserts that a conversation/question to be responded to can be read/heard by a user to be converted to phonemes via mental process as previously disclosed. The contextual setting being in a conversation does not prevent the mental process interpretation.
All of these steps can be performed in the mind and/or using pen and paper. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea (Step 2A, Prong one, Yes).
This judicial exception is not integrated into a practical application because the addition of generically recited computer elements does not add a meaningful limitation to the abstract idea because they amount to simply implementing the abstract idea on a computer. The claims are directed to an abstract idea. The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception ( Step 2A, Prong two, No). As discussed above, with respect to integration of an abstract idea into a practical application, the additional element of “converting”, “training”, “receiving”, “executing” are merely for the purpose of data gathering, storing, processing, and/or insignificant extra-solution activity that amount to no more than mere instructions to apply the exception using a generic computer component. Fig. 16 of the instant application disclose(s) applying the method to a generic computing device such as a PC (computer 970). Mere instructions to apply the exception using a generic computer component cannot provide an inventive concept. Therefore, the claims are not patent eligible (Step 2B, No).
Similarly, dependent claim(s) 2-5 include additional steps that are considered “insignificant extra-solution activity to judicial exception” because they fail to provide meaningful significance that goes beyond generally linking the use of an abstract idea to a particular technological environment.
For example, claim 2 reads on training including training data which combines the text and phonetic letters output from the text-to-phoneme conversion and pre-training the model using this data. As previously disclosed, text and pronunciations of said text can be written together on paper. This writing can then be presented to someone, i.e. one not familiar with the language, to be spoken as a means of pre-training. Applying a language model to perform this operations is applying a generic computing component to a mental process.
Claim 3 reads on adding noise to a sequence of phonetic letters to generate a noise-added sequence of phonetic letters, generating fine-tuning training data with the noise-added sequence, and fine-tuning the language model using this data. A person writing a phonetic transcript can add additional phonemes and/or sounds via writing the same way the original phoneme sequence was written. Fine-tuning a language model using this data is applying a generic computing component to a mental process. The fine-tuning sequence can be provided to someone with an understanding of the language, so they can identify unexpected phonemes.
Claim 4 reads on the same operations as claim 3 but for a “pre-trained language model”. As previously disclosed, the operations of claim 3 are a mental process. Applying a mental process to a “pre-trained language model” is limiting the mental process to a generic computing environment. This does not provide an inventive step. Claim 5 reads on the same operations as claim 3 and 4 but for the recitation of “additional training data” for “additionally pre-training said pre-trained language model”. Extending the previously disclosed rationale as to why the operations are mental processes, defining the training data to now be “additional” does not prevent the previous rationale from still being applicable.
Therefore, these claims are also not patent eligible.
Claim 7 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claim(s) does/do not fall within at least one of the four categories of patent eligible subject matter because the claim is directed to a “trained language model generated by machine learning” with no physical components recited. This indicates a model which does not have a physical or tangible form such as a computer program per se (otherwise referred to as “software per se”) when claimed as a product without any structural limitations (see MPEP 2106.03, Section I).
Claim Rejections - 35 USC § 102
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.
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claim(s) 1, 4-7 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Arel et al. (US-10210861-B1), hereinafter Arel.
Regarding claim 1, Arel discloses: A language model training device ([Col. 3, Lines 54-55] acoustic model may be trained, [An acoustic model which receives audio data resulting in phonemes (see Fig. 1) indicates the acoustic model to be a language model]), comprising:
a converting means for converting natural language text to output a sequence of phonetic letters ([Fig. 2A, Grapheme To Phoneme Converter 228], [Col. 4, Lines 15-25] For each synthetic training data item, a grapheme to phoneme converter may convert the textual representation of the synthetic sentence into a sequence of phonemes that represent the synthetic sentence); and
a training means for training a language model using said text and said sequence of phonetic letters output from said converting means ([Fig. 2A, Modified Training Dataset 230 comprised of Sequence of Phonemes 235], [Col. 7, Lines 55-61] the text transcriptions 215 of the training data items 205 in the initial training dataset 202 are input into a grapheme to phoneme converter 228 that converts the text transcriptions 215 into sequences of phonemes 235, [Generating a training dataset, i.e. training means, for training an acoustic, i.e. language, model, wherein the phoneme sequence is generated based on text, indicates the training means to be based on the text and phoneme sequence as the text is required for generating the phonemes as previously disclosed]).
Regarding claim 4, Arel discloses: the language model training device according to claim 1.
Arel further discloses:
wherein said language model includes a pre-trained language model ([Fig. 2A, Initial Training Dataset 202 and Modified Training Data Items 232 comprised of Audio Data 210 and/or Distorted Audio Data 224], [Considering at least two training sets with an original and modified dataset, indicates the training using the original dataset to be a pre-training as compared to the training with a modified dataset]); and
said training means includes:
a noise adding means for adding noise to said sequence of phonetic letters to generate a noise-added sequence of phonetic letters ([Fig. 2B, Phoneme Distorter 288], [Col. 10, Lines 44-52] training data items 280 of the modified synthetic training dataset 275 may be input into a phoneme distorter 288 prior to being input into the transcoder 120 for training. The phoneme distorter 228 may include a number of rules that perform operations such as inserting one or more additional phonemes to a sequence of phonemes 285, deleting one or more phonemes to a sequence of phonemes 285, and/or substituting one or more phonemes from the sequence of phonemes 285);
a training data forming means for forming training data for fine-tuning said language model pre-trained by said pre-training means, using said text, said sequence of phonetic letters and said noise-added sequence of phonetic letters ([Fig. 2B, Distorted Synthetic Training Dataset 290], [Col. 10, Lines 65-67]-[Col. 11, Line 1] The distorted synthetic training dataset 290 may be input into the transcoder 120 during training along with or instead of the modified synthetic training dataset 275, [Fig. 5, Estimate and Update Acoustic Model Parameters using Modified data item S515], [Wherein the distorted training dataset is generated after the initial and/or modified training sets, indicating the distorted set to be for fine-tuning with respect to the earlier training datasets. Further, if the modified and distorted training sets are used together, as disclosed above, this indicates training the acoustic model using the sequence of phonetic letters and noise-added phonetic letters, wherein the phonetic sequences are necessarily based on text via the grapheme-to-phoneme conversion. Further, wherein the modification of Fig. 5 is distortion (see S512) which tracks to noise-addition]);
a fine-tuning means for fine-tuning said pre-trained said language model by using said training data ([As previously disclosed, training an acoustic model with modified training data items (as disclosed in Fig. 2A), wherein that model can also be trained on distorted synthetic training data items (as disclosed in Fig. 2B), indicates the additional training with distorted synthetic data to be “fine-tuning” with respect to the earlier training operations with a more limited type of data]).
Regarding claim 5, Arel discloses: the language model training device according to claim 1.
Arel further discloses:
wherein said language model includes a pre-trained language model ([Fig. 2A, Initial Training Dataset 202 and Modified Training Data Items 232 comprised of Audio Data 210 and/or Distorted Audio Data 224], [Considering at least two training sets with an original and modified dataset, indicates the training using the original dataset to be a pre-training as compared to the training with a modified dataset]); and
said training means includes:
a noise adding means for adding noise to said sequence of phonetic letters to generate a noise-added sequence of phonetic letters ([Fig. 2B, Phoneme Distorter 288], [Col. 10, Lines 44-52] training data items 280 of the modified synthetic training dataset 275 may be input into a phoneme distorter 288 prior to being input into the transcoder 120 for training. The phoneme distorter 228 may include a number of rules that perform operations such as inserting one or more additional phonemes to a sequence of phonemes 285, deleting one or more phonemes to a sequence of phonemes 285, and/or substituting one or more phonemes from the sequence of phonemes 285);
an additional training data forming means for forming additional training data for additionally fine-tuning said language model pre-trained by said pre-training means, using said text, said sequence of phonetic letters and said noise-added sequence of phonetic letters ([Fig. 2B, Distorted Synthetic Training Dataset 290], [Col. 10, Lines 65-67]-[Col. 11, Line 1] The distorted synthetic training dataset 290 may be input into the transcoder 120 during training along with or instead of the modified synthetic training dataset 275, [Fig. 5, Estimate and Update Acoustic Model Parameters using Modified data item S515], [Wherein the distorted training dataset is generated after the initial and/or modified training sets, indicating the distorted set to be for fine-tuning with respect to the earlier training datasets. Further, if the modified and distorted training sets are used together, as disclosed above, this indicates training the acoustic model using the sequence of phonetic letters and noise-added phonetic letters, wherein the phonetic sequences are necessarily based on text via the grapheme-to-phoneme conversion. Further, wherein the modification of Fig. 5 is distortion (see S512) which tracks to noise-addition. Citing an “additional” training/fine-tuning when there is no first operations indicates the additional operation to be a first.]);
an additional fine-tuning means for additionally fine-tuning said pre-trained said language model by using said training data ([As previously disclosed, training an acoustic model with modified training data items (as disclosed in Fig. 2A), wherein that model can also be trained on distorted synthetic training data items (as disclosed in Fig. 2B), indicates the additional training with distorted synthetic data to be “fine-tuning” with respect to the earlier training operations with a more limited type of data. Citing an “additional” training/fine-tuning when there is no first operations indicates the additional operation to be a first.]).
Regarding claim 6, Arel discloses: a dialogue device realizing speech-based dialogue with a user ([Col. 2, Lines 50-55] users are able to speak freely to the conversational agent), comprising:
a trained language model generated by machine learning ([Fig. 2A, Modified Training Dataset 230 comprised of Sequence of Phonemes 235], [Col. 7, Lines 55-61] the text transcriptions 215 of the training data items 205 in the initial training dataset 202 are input into a grapheme to phoneme converter 228 that converts the text transcriptions 215 into sequences of phonemes 235, [Generating a training dataset, i.e. training means, for training an acoustic, i.e. language, model, wherein the phoneme sequence is generated based on text, indicates the training means to be based on the text and phoneme sequence as the text is required for generating the phonemes as previously disclosed]) using at least natural language text and a sequence of phonetic letters obtained by converting the text ([Fig. 2A, Grapheme To Phoneme Converter 228], [Col. 4, Lines 15-25] For each synthetic training data item, a grapheme to phoneme converter may convert the textual representation of the synthetic sentence into a sequence of phonemes that represent the synthetic sentence);
a semantic interpretation module with said trained language model ([Fig. 6, S610 Process the new utterance]), for receiving as an input speech information of said user ([Fig. 6, S605 Receive a new utterance at a first machine learning model trained as an acoustic model]); and
an utterance/response module for receiving as an input the speech information of said user and for executing a dialogue with the user under control of said semantic interpretation module ([Col. 2, Lines 45-50] conversational agent may be trained to understand and respond accurately to every way that a customer might order off of a restaurant's menu, [The examiner asserts that, as the semantic interpretation module is currently claimed to be “for receiving…speech information of said user”, a speaking user will necessarily be in control of what is spoken and received by the semantic interpretation module]).
Regarding claim 7, Arel discloses: a trained language model generated by machine learning ([Fig. 2A, Modified Training Dataset 230 comprised of Sequence of Phonemes 235], [Col. 7, Lines 55-61] the text transcriptions 215 of the training data items 205 in the initial training dataset 202 are input into a grapheme to phoneme converter 228 that converts the text transcriptions 215 into sequences of phonemes 235, [Generating a training dataset, i.e. training means, for training an acoustic, i.e. language, model, wherein the phoneme sequence is generated based on text, indicates the training means to be based on the text and phoneme sequence as the text is required for generating the phonemes as previously disclosed]), using at least natural language text and a sequence of phonetic letters obtained by converting the text ([Fig. 2A, Grapheme To Phoneme Converter 228], [Col. 4, Lines 15-25] For each synthetic training data item, a grapheme to phoneme converter may convert the textual representation of the synthetic sentence into a sequence of phonemes that represent the synthetic sentence).
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 2-3 is/are rejected under 35 U.S.C. 103 as being unpatentable over Arel in view of Gupta et al. (US-8438005-B1), hereinafter Gupta.
Regarding claim 2, Arel discloses: the language model training device according to claim 1.
Arel does not disclose:
training data forming means for forming training data for training said language model by combining said text and the sequence of phonetic letters output from said converting means.
Gupta discloses:
training data forming means for forming training data for training said language model by combining said text and the sequence of phonetic letters output from said converting means ([Fig. 1, Comparison Engine 150 receiving Phonetic Representation 145 and Character Combination 135], [Wherein the character combination consists of English letters, indicating it to be natural language text. Further, this data is taken in view of the previously disclosed training operation of Arel, wherein the data of Gupta could be used as the training data of Arel as Arel discloses training data of phoneme sequences generated from text as would be compared to the gathered phoneme sequences of Gupta for comparison. The comparison of Gupta has the same functionality as the transcoder training accuracy of Arel (see [Col. 11, Lines 65-67] and [Col. 12, Lines 1-5] of Arel for discussion of the transcoder]).
Arel and Gupta are considered analogous art within existing phoneme sequence modification for learning purposes. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Arel to incorporate the teachings of Gupta, because of the novel way to compare phonetic representations of Indic language words to English transcriptions of previously correctly identified Indic words, improving the accuracy of English transcriptions of new Indic words (Gupta, [Col. 3, Lines 5-20]).
Arel further discloses:
a pre-training means for pre-training said language model using said training data ([Fig. 2A, Initial Training Dataset 202 and Modified Training Data Items 232 comprised of Audio Data 210 and/or Distorted Audio Data 224], [Considering at least two training sets with an original and modified dataset, indicates the training using the original dataset to be a pre-training as compared to the training with a modified dataset]).
Regarding claim 3, Arel in view of Gupta discloses: the language model training device according to claim 2.
Arel further discloses:
a noise adding means for adding noise to said sequence of phonetic letters to generate a noise-added sequence of phonetic letters ([Fig. 2B, Phoneme Distorter 288], [Col. 10, Lines 44-52] training data items 280 of the modified synthetic training dataset 275 may be input into a phoneme distorter 288 prior to being input into the transcoder 120 for training. The phoneme distorter 228 may include a number of rules that perform operations such as inserting one or more additional phonemes to a sequence of phonemes 285, deleting one or more phonemes to a sequence of phonemes 285, and/or substituting one or more phonemes from the sequence of phonemes 285);
a training data forming means for forming training data for fine-tuning said language model pre-trained by said pre-training means, using said text, said sequence of phonetic letters and said noise-added sequence of phonetic letters ([Fig. 2B, Distorted Synthetic Training Dataset 290], [Col. 10, Lines 65-67]-[Col. 11, Line 1] The distorted synthetic training dataset 290 may be input into the transcoder 120 during training along with or instead of the modified synthetic training dataset 275, [Fig. 5, Estimate and Update Acoustic Model Parameters using Modified data item S515], [Wherein the distorted training dataset is generated after the initial and/or modified training sets, indicating the distorted set to be for fine-tuning with respect to the earlier training datasets. Further, if the modified and distorted training sets are used together, as disclosed above, this indicates training the acoustic model using the sequence of phonetic letters and noise-added phonetic letters, wherein the phonetic sequences are necessarily based on text via the grapheme-to-phoneme conversion. Further, wherein the modification of Fig. 5 is distortion (see S512) which tracks to noise-addition]);
a fine-tuning means for fine-tuning said pre-trained said language model by using said training data ([As previously disclosed, training an acoustic model with modified training data items (as disclosed in Fig. 2A), wherein that model can also be trained on distorted synthetic training data items (as disclosed in Fig. 2B), indicates the additional training with distorted synthetic data to be “fine-tuning” with respect to the earlier training operations with a more limited type of data]).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Tan et al. (US-20240233706-A1) discloses “According to implementations of the subject matter described herein, a solution is proposed for text to speech. In this solution, an initial phoneme sequence corresponding to text is generated, the initial phoneme sequence comprising feature representations of a plurality of phonemes. A first phoneme sequence is generated by inserting a feature representation of an additional phoneme into the initial phoneme sequence, the additional phoneme being related to a characteristic of spontaneous speech. The duration of a phoneme among the plurality of phonemes and the additional phoneme is determined by using an expert model corresponding to the phoneme, and a second phoneme sequence is generated based on the first phoneme sequence. Spontaneous-style speech corresponding to the text is determined based on the second phoneme sequence. In this way, spontaneous-style speech with more varying rhythms can be generated based on spontaneous-style additional phonemes and multiple expert models” (abstract). See entire document.
Sivadas (US-20080126093-A1) discloses “An apparatus for providing a language based interactive multimedia system includes a selection element, a comparison element and a processing element. The selection element may be configured to select a phoneme graph based on a type of speech processing associated with an input sequence of phonemes. The comparison element may be configured to compare the input sequence of phonemes to the selected phoneme graph. The processing element may be in communication with the comparison element and configured to process the input sequence of phonemes based on the comparison” (abstract). See entire document.
Gao et al. (US-9697201-B2) discloses “A speech-to-speech (S2S) translation system may utilize a damaging channel model to adapt machine translation (MT) training data so that a MT engine of the S2S translation system that is trained with the adapted training data can make better use of output received from an automated speech recognition (ASR) engine of the S2S translation system. The S2S translation system may include a MT training module that uses MT technology in order to simulate a particular ASR engine output by treating the ASR engine as a “noisy channel”. A process may include modeling ASR errors of a particular ASR engine based at least in part on output of the ASR engine to create an ASR simulation model, and performing machine translation to generate training data based at least in part on the ASR simulation model. The MT engine of the S2S translation system may then be trained using the generated training data” (abstract). See entire document.
Sankoda et al. (US-20210005204-A1) discloses “A method for transcription is performed by a computer. The method includes: accepting input of a voice after causing a display unit to display a sentence including a plurality of words; acquiring first sound information being information concerning sounds corresponding to the sentence; acquiring second sound information being information concerning sounds of the voice accepted in the accepting; specifying a portion in the first sound information having a prescribed similarity to the second sound information; and correcting a character string in the sentence corresponding to the specified portion based on a character string corresponding to the second sound information” (abstract). See entire document.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to THEODORE JOHN WITHEY whose telephone number is (703)756-1754. The examiner can normally be reached Monday - Friday, 8am-5pm.
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, Andrew Flanders can be reached at (571) 272-7516. 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.
/THEODORE WITHEY/Examiner, Art Unit 2655
/ANDREW C FLANDERS/Supervisory Patent Examiner, Art Unit 2655
Prosecution Insights