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 .
Response to Amendment
2. In response to the office action mailed on 01/16/2026, applicant filed an amendment on 04/06/2026, amending claims 1, 7-11, 14, 15, 21, and 22. Claims 12-13 were previously cancelled. The pending claims are 1-11 and 14-22.
Response to Arguments
3. Applicant's arguments filed 04/06/206 have been fully considered but they are not persuasive.
With regard to 35 U.S.C. 101, applicant argues that amended claim 1 is not directed to any abstract concept or mental process that can be performed in the human mind and amended claim 1 as a whole includes additional elements that are sufficient to amount to significantly more than the alleged judicial exception because speech-to-speech translation samples are expanded, and the model training precision can be improved.
The examiner notes, even when viewed in combination, the additional elements do not integrate the recited judicial exception into a practical application. The claimed steps obtaining a speech recognition sample and a real speech-to-speech translation sample; generating a pseudo-labeled speech-to-speech translation sample based on the speech recognition sample; and training a speech-to-speech translation model based on the pseudo-labeled speech- to-speech translation sample and the real speech-to-speech translation sample, wherein the pseudo-labeled speech-to-speech translation sample is constructed based on a source-language speech feature of the speech recognition sample, are recited at high level of generality. They are not indicative of an inventive concept. The addition of insignificant extra-solution activity does not amount to an inventive concept, particularly when the activity is well-understood or conventional. According to MPEP 2106.05(f), the recitation of claim limitations that attempt to cover any solution to an identified problem with no restriction on how the result is accomplished and no description of the mechanism for accomplishing the result, does not integrate a judicial exception into a practical application or provide significantly more.
With regard to prior art, applicant argues that the prior art Wang does not teach wherein the pseudo-labeled speech-to-speech translation sample is constructed based on a source-language speech feature of the speech recognition sample. Applicant asserts that Wang describes a pseudo-label generation feature, which is intended to generate supervisory signals for the self-supervised pre-training of automatic speech recognition (ASR) models and address the issue that ASR model training relies on a large volume of manually labeled speech data. However, Wang only discloses how to perform alignment processing on existing speech and text to generate acoustic labels, and does not generate any new speech data sample, new paired data samples in different languages.
In response to applicant's argument that the references fail to show certain features of the invention, it is noted that the features upon which applicant relies (i.e., generate any new speech data sample, new paired data samples in different languages) are not recited in the rejected claim(s). Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993).
Claim 1 recites “generating a pseudo-labeled speech-to-speech translation sample”
The prior art Wang teaches, at Fig. 2 and [0045]- [0046], a process that generates pseudo-labels 216. The pseudo-labels 216 correspond to speech utterances in the unlabeled data 208. The unlabeled speech data, along with labeled data which comprise speech data and corresponding speech transcription data, can be synthesized speech data, processed speech data, and/or raw speech data. For raw speech data, natural language audio is obtained from a plurality of locations and applications. In some instances, natural language audio is extracted from previously recorded files such as video recordings having audio or audio-only recordings. Some examples of recordings include videos, podcasts, voicemails, voice memos, songs, etc. ([0030]). Raw speech data such as audio recordings, voicemails, voice memos, songs, from which the pseudo-labeled speech-to-speech translation sample, is constructed is necessarily a source-language speech feature of the speech recognition sample.
As per the rest of the claims, and combinations of prior art reference, applicant has no further arguments beside the ones mentioned above. Therefore, all the combinations of prior art reference mentioned above are valid, and all other claims are rejected for the same reasons as set above.
Claim Rejections - 35 USC § 101
4. 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.
Claims 1-11 and 14-22 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Step 1: Is the claimed invention to a process, machine, manufacture or composition of matter?
The claimed invention, at independent claims 1, 14, and 15, is directed to a method (process), system (machine), and computer readable medium (manufacture) for obtaining a speech recognition sample and a real speech-to-speech translation sample; generating a pseudo-labeled speech-to-speech translation sample based on the speech recognition sample; and training a speech-to-speech translation model based on the pseudo-labeled speech-to- speech translation sample and the real speech-to-speech translation sample.
Step 2A, prong 1: Does the claim recite an abstract idea, law or nature, or natural phenomenon?
Under the 35 U.S.C. 101 new guidelines, the broadest reasonable interpretation of the claims, the claimed steps fall within the “Mental Processes” grouping of abstract ideas because they cover concepts performed in the human mind, including observation, evaluation, judgment, and opinion. See MPEP 2106.04(a)(2), subsection III.
The steps of obtaining speech recognition sample and a real speech-to-speech translation sample; generating a pseudo-labeled speech-to-speech translation sample based on the speech recognition sample, could be practically performed in the human mind using observation, evaluation, judgment, and opinion. For example, a human can receive the speech recognition sample and real speech-to-speech translation sample, and generate corresponding labels or tags based on the received sample without using a machine. As to the step of training a speech-to-speech translation model based on the pseudo-labeled speech-to- speech translation sample and the real speech-to-speech translation sample, it encompasses mental processes practically performed in the human mind primarily by creating labeled datasets, fine tuning and evaluating performance via observation, evaluation, judgment, and opinion. See MPEP 2106.04(a)(2), subsection III.
Therefore, the claimed steps fall within the mental process grouping of abstract ideas
Step 2A, prong 2: Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements of “a processor and a memory, wherein the memory is configured to store a computer program, and the processor is configured to invoke and run the computer program stored in the memory”. The processor is recited at a high level of generality, and it amounts to no more than mere instructions to apply the exception using a generic computer. See MPEP 2106.05(f). The claims do not recite a specific, unconventional machine or physical transformation beyond processing of data (speech/text features) and application of generic computing resources. The transformation of data (speech feature, target-language speech feature) is an information/content manipulation that, without more, is an abstract idea. Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application, and the claims are directed to the judicial exception.
Step 2B: Does the claim recite additional elements that amount to significantly more than the abstract idea?
As to whether the claims as a whole amount to significantly more than the recited exception, i.e., whether any additional element, or combination of additional elements, adds an inventive concept to the claim (Step 2B), as explained above in Step 2A, Prong 2, the use of “computer”, “processor” is at high level of generality, and even when considered in combination, these additional elements represent mere instructions to apply an exception and insignificant extra-solution activity, and therefore do not provide an inventive concept. Accordingly, the claims are ineligible.
The dependent claims further refer and describe the process of pre-training/fine-tuning (claims 2–3), labeling (claim 4), up-sampling (claim 5), specific data elements (claim 6) describe well-known training techniques and data representations, translating from one language to the other (claim 7). These steps, as claimed, are generic uses of routine machine learning techniques and do not amount to an integration of the abstract idea into a specific practical application. Claims 8-10 recite a neural architecture and provide structural detail but largely recite conventional components (encoder, attention modules, decoder modules, transformer components) and general data-flow descriptions. The presence of attention modules and transformer-like components, without a clear claim of a specific non-conventional arrangement that provides a concrete technical improvement, is insufficient to transform the abstract idea into a patent-eligible application. Claim 11 relate to collecting speech data, which is considered a mere data gathering at a high level of generality, and thus are insignificant extra-solution activity. Claims 16-22 recite similar steps as in claims 2-11, which encompasses a mental process that is practically performed in the human mind, as explained above. Accordingly, claims 1-11 and 14-22 are directed to an abstract idea, and are not patent eligible.
Claim Rejections - 35 USC § 103
5. 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.
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.
Claims 1-4, 6, 11, 14-18, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Wang (US 2025/0157459) in view of Matusov (US 20200226327).
As per claim 1, Wang teaches obtaining a speech recognition sample and a real speech-to-speech translation sample ([0030], wherein sets of unlabeled data which comprise unlabeled speech data and sets of labeled data which comprise speech data and corresponding speech transcription data (e.g., text data) which are labels for the speech data are obtained as training data. In either type of set of training data, the speech data can be synthesized speech data, processed speech data, and/or raw speech data. [0031]- [0032], the real speech data is extracted from actively streaming content which is live continuous speech such as a news broadcast, phone call, virtual or in-person meeting, etc., and comprises one or more spoken languages of the world's spoken languages. The speech data can be also post-processed synthesized speech data or raw speech data that has been subsequently processed and/or filtered);
generating a pseudo-labeled speech-to-speech translation sample based on the speech recognition sample ([0062]- [0065], generating sets of pseudo-labels associated with the unlabeled speech data by applying a hybrid automatic speech recognition model to both the set of decoded word sequences and the set of unlabeled speech data); and
training a speech-to-speech translation model based on the pseudo-labeled speech-to- speech translation sample and the real speech-to-speech translation sample ([0042], wherein said training engine 153 is configured to train the ASR model and the speech processing model with pseudo-labeled training data; also fine-tune the speech processing model with labeled data. See also, [0071], wherein said, although not shown in FIG. 8, the disclosed methods also include using pseudo-labels generated with the aforementioned novel techniques to pretrain or refine training of a speech processing model for improving accuracy and/or efficiency of the speech processing model in performing a speech processing task (e.g., ASR, speaker recognition, speech separation tasks, transcription processing, language translation, or any other speech processing tasks)).
Wang teaches the training speech data can be synthesized speech data, processed speech data, and/or raw speech data and performing speech processing tasks (e.g., ASR, speaker recognition, speech separation tasks, transcription processing, language translation, or any other speech processing tasks) ([0030], [0071]). Wang may not explicitly disclose speech-to-speech translation. Matusov in the same field of endeavor teaches performing multi-lingual speech-to-speech translation in an end-to-end workflow ([0017]). Therefore, it would have been obvious at the time the application was filed to use Matusov’ s multi-lingual speech-to-speech translation feature with the system of Wang, in order to facilitate real-time communication across different languages by capturing spoken words and playing them back as translated audio.
As to wherein the pseudo-labeled speech-to-speech translation sample is constructed based on a source-language speech feature of the speech recognition sample, the prior art Wang teaches, at Fig. 2 and [0045]- [0046], a process that generates pseudo-labels 216. The pseudo-labels 216 correspond to speech utterances in the unlabeled data 208. The unlabeled speech data, along with labeled data which comprise speech data and corresponding speech transcription data, can be synthesized speech data, processed speech data, and/or raw speech data. For raw speech data, natural language audio is obtained from a plurality of locations and applications. In some instances, natural language audio is extracted from previously recorded files such as video recordings having audio or audio-only recordings. Some examples of recordings include videos, podcasts, voicemails, voice memos, songs, etc. ([0030]- [0032]). Raw speech data such as audio recordings, voicemails, voice memos, songs, phone calls, live continuous speech, from which the pseudo-labeled speech-to-speech translation sample, is constructed is necessarily a source-language speech feature of the speech recognition sample.
As per claim 2, Wang teaches wherein the training a speech-to-speech translation model based on the pseudo-labeled speech-to-speech translation sample and the real speech-to-speech translation sample comprises: pre-training the speech-to-speech translation model based on the pseudo-labeled speech- to-speech translation sample, and fine-tuning the pre-trained speech-to-speech translation model based on the real speech-to-speech translation sample ([0025], generating a pretrained speech processing model by applying the pseudo labeled training data to the speech processing model; and [0046], [0071], wherein said, the pre-trained speech processing model can then be further trained (e.g., fine-tuned) using labeled data).
As per claim 3, Wang teaches fine-tuning the pre-trained speech-to-speech translation model by using the real speech-to- speech translation sample; or fine-tuning the pre-trained speech-to-speech translation model based on the real speech-to- speech translation sample and the pseudo-labeled speech-to-speech translation sample ([0046], the labeled data used to fine-tune the pre-trained speech processing model is the same set or sub-set of the same set of labeled data used to train the speech processing model as in [0031], [0032], and [0042]).
As per claim 4, Wang teaches using labeled data, unlabeled data, and generating a first set of pseudo-labeled data and second set of pseudo-labeled data ([0060]- [0065]). Wang may not explicitly recite labeling the real speech-to-speech translation sample with a first label, wherein the first label is used for identifying the real speech-to-speech translation sample as a real sample; and labeling the pseudo-labeled speech-to-speech translation sample with a second label, wherein the second label is used for identifying the pseudo-labeled speech-to-speech translation sample as a pseudo-labeled sample, as claimed. However, it’s common to label different training sets with different labels to provide distinctive sets, which obviously is the role of labels as it’s well known in the art. Therefore, it would have been obvious a the time the application was filed foe the system of Wang to label the real speech-to-speech translation sample with a first label, wherein the first label is used for identifying the real speech-to-speech translation sample as a real sample; and label the pseudo-labeled speech-to-speech translation sample with a second label, wherein the second label is used for identifying the pseudo-labeled speech-to-speech translation sample as a pseudo-labeled sample. This would provide organized reference data and accurate results.
As per claim 6, Wang teaches wherein the real speech-to-speech translation sample comprises: a first source-language speech feature, first source-language text, a first target-language speech feature, and first target-language text; and the speech recognition sample comprises: a second source-language speech feature and second source- language text ([0031]- [0033], [0071], wherein speech training data comprises spoken language utterances from a plurality of sources, including applications, meetings comprising one or more speakers, ambient environments including background noise and human speakers speaking one or more spoken languages of the world's spoken languages; and the sets of labeled training data comprise transcription data and natural language audio and/or simulated audio that comprises speech utterances corresponding to words, phrases, and sentences included in the text data).
As per claim 11, obtaining a source-language speech feature; and inputting the source-language speech feature into a speech-to-speech translation model to obtain a target-language speech feature corresponding to the source-language speech feature ([0030]- [0031] and [0060]- [0063]). Wang may not explicitly disclose speech-to-speech translation. Matusov in the same field of endeavor teaches performing multi-lingual speech-to-speech translation in an end-to-end workflow ([0017]). Therefore, it would have been obvious at the time the application was filed to use Matusov’ s multi-lingual speech-to-speech translation feature with the system of Wang, in order to facilitate real-time communication across different languages by capturing spoken words and playing them back as translated audio.
As per claims 14, 16-18 and 20, system claims 14, 16-18 and 20 and method claims 1-4 and 6 are related as apparatus and the method of using same, with each claimed element's function corresponding to the claimed method step. Accordingly claims 14, 16-18, 20 are similarly rejected under the same rationale as applied above with respect to method claims 1-4, 6. Furthermore, Wang teaches one or more processors; and memory storing thereon instructions, as claimed (Fig. 1).
As per claim 15, Wang teaches a computer readable medium ([0073]. The remaining steps are rejected under the same rationale as applied to the method steps of rejected claim 1.
Claims 5, 7, 19, and 21 are rejected under 35 U.S.C. 103 as being unpatentable over Wang (US 2025/0157459) in view of Matusov (US 20200226327), and further in view of Gao (US 20210020161).
As per claims 5 and 19, Wang teaches fine-tuning the pre-trained speech-to-speech translation model based on the real speech-to-speech translation sample and the pseudo-labeled speech-to-speech translation sample as evidenced by paragraphs [0025], and [0046], [0071]. Wang may not explicitly disclose performing up-sampling on the real speech-to-speech translation sample to obtain an up- sampled speech-to-speech translation sample; and fine-tuning the pre-trained speech-to-speech translation model by using the up-sampled speech-to-speech translation sample and the pseudo-labeled speech-to-speech translation sample. However, upsampling training data is well known in the art as evidenced by Gao. Gao in the same field of endeavor teaches training a speech translation processing system, wherein an up-sampling process may be performed at the Vocoder to increase the number of output audio frames ([0016]). Therefore, it would have been obvious a the time the application was filed to use the upsampling feature of Gao with the system of Wang, in order to perform up-sampling on the real speech-to-speech translation sample to obtain an up- sampled speech-to-speech translation sample, and fine-tuning the pre-trained speech-to-speech translation model by using the up-sampled speech-to-speech translation sample and the pseudo-labeled speech-to-speech translation sample, as claimed. This would improve model performance and provide cost-effective procedures.
As per claims 7 and 21, Wang teaches wherein the generating a pseudo-labeled speech-to-speech translation sample based on the speech recognition sample ([0062]- [0065], generating sets of pseudo-labels associated with the unlabeled speech data by applying a hybrid automatic speech recognition model to both the set of decoded word sequences and the set of unlabeled speech data). Wang may not explicitly disclose translating the second source-language text to obtain second target-language text; and performing synthesis on the second target-language text to obtain a second target-language speech feature, wherein the pseudo-labeled speech-to-speech translation sample comprises: the second source-language speech feature, the second source-language text, the second target-language text, and the second target-language speech feature. Gao in the same field of endeavor teaches at Fig. 2(a) a source language text is input to a text-to-text translation module, producing output text in the target language. Then, the text in the target language is input into a text-to-speech module to obtain a second target-language speech feature ([0122]- [0124]). Therefore, it would have been obvious a the time the application was filed to use the above features of Gao with the system of Wang, in order to improve user experience by making content accessible in native languages and via different modes.
As per claim 8-10 and 22, the prior art does not teach wherein the speech-to-speech translation model comprises: an encoder module, a first attention module, a first decoder module, N second attention modules, and N second decoder modules, wherein N is a positive integer, and the N second attention modules are in a one-to-one correspondence with the N second decoder modules; the encoder module is configured to obtain a source-language speech feature, and process the source-language speech feature to obtain a plurality of groups of first hidden-state representations corresponding to the source-language speech feature; the first attention module is configured to obtain one group of the plurality of groups of first hidden-state representations, and a first vector corresponding to each time step that is output by the first decoder, and process the group of first hidden-state representations and the first vector corresponding to each time step, to obtain a first attention representation corresponding to each time step; the first decoder module is configured to obtain a second vector corresponding to each time step, process the second vector corresponding to each time step to obtain the first vector corresponding to each time step, output the first vector corresponding to each time step to the first attention module, obtain the first attention representation corresponding to each time step, and process the first attention representation corresponding to each time step to obtain a target- language speech feature corresponding to the source-language speech feature; in a training stage of the speech-to-speech translation model, each of the second attention modules is configured to obtain one group of the plurality of groups of first hidden-state representations, and a third vector corresponding to each time step that is output by a second decoder corresponding to the second attention module, and process the group of first hidden-state representations and the third vector corresponding to each time step, to obtain a second attention representation corresponding to each time step; and the second decoder module corresponding to the second attention module is configured to obtain a fourth vector corresponding to each time step, process the fourth vector corresponding to each time step to obtain the third vector corresponding to each time step, output the third vector corresponding to each time step to the second attention module, obtain the second attention representation corresponding to each time step, and process the second attention representation corresponding to each time step, to obtain a secondary representation corresponding to the source- language speech feature.
Conclusion
6. The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. See PTO-892.
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ABDELALI SERROU whose telephone number is (571)272-7638. The examiner can normally be reached M-F 9 Am - 5 PM.
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-Louis 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.
/ABDELALI SERROU/Primary Examiner, Art Unit 2659
06/09/2026
Prosecution Insights