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 .
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.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claims recite a mental process (abstract idea), that can be performed by a human mind/using a pen and paper:
As per independent claims 1,19, 20, the steps of obtaining a dataset containing speech features of a target speaker, the steps of obtaining a data set containing speech features of reference speakers, and determining the size of the dataset being larger than the size of the target speaker datasets, taking a subset of the speech features of the reference speakers, training an acoustic model to produce synthesized reference speech features based on the trained acoustic model; and training the post processing model based on synthesized speech features, from the reference speaker dataset and the speech features included in the target speaker data set;
all can be performed by a human mind/using a pen and paper. For example, a human can determine the type of amplitude and pitch of a target speech, a human can analyze voice files for amplitude and pitch and syllable duration (which has more features, bigger feature size than the reference speech), choosing/taking/selecting certain pitch/duration/amplitude from different reference speakers, combining the different source elements, and use it as features for the synthesized target speech.
As per claims 2-4, a user can choose random samples of speech features.
As per claims 5-6, a user can choose any ratio of features of the reference speakers to the target speaker.
As per claims 7-8, 18, a user can choose prosody characteristics.
Claims 9-10, 15, 17 are toward the relationship between the data sets.
Claims 11-14 are descriptive-type claims, describing certain elements of claim 1, without further modifications.
Claim 16 claims the well known, routine, step of synthesizing speech. In this instant, Desjardins applies – see MPEP 2106.04 (d)(1)(2). However, in further review of the claims and applicants specification, claim 16 does not pass the Step 2A, Prong 2 test. See rationale provided below.
This judicial exception is not integrated into a practical application because the claim features are toward re-organizing/selecting certain speech features. The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Furthermore, as per the independent claims (and dependent claim 16), the abstract idea of mental steps is not modified into a practical application (STEP 2A, PRONG 2, not satisfied). Furthermore – Ex Parte Desjardins, MPEP2106.04 (d) (1), 2nd/last paragraphs, are not satisfied, wherein the claimed steps are not explained, in detail, in improving the art. Step 2A EXCLUDES consideration of whether the additional elements represent well-understood, routine, conventional activity, IF an explicit explanation of an improvement in the art is directly tied to the detailed claim features. At this juncture, no such detailed connection has been found.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claim 19 is rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being incomplete for omitting essential structural cooperative relationships of elements, such omission amounting to a gap between the necessary structural connections. See MPEP § 2172.01. The omitted structural cooperative relationships are: how the claimed processor executes with/side/parallel to the claimed vocoder, in claim 1, and the generation of signals (eventually found in claim 16).
Allowable Subject Matter
Claims 1-20 are allowable over the prior art of record, and would be allowed once amended to overcome the 35 USC 101 and 35 USC 112 rejections, as applied above.
The following is a statement of reasons for the indication of allowable subject matter:
As per the independent claims, what is notoriously old and well known in the prior art, are various subfeatures found in the claims. As to the general concept of using an expanded vector/feature space from reference speakers, and down to size of a target speaker, see as a representative prior art, Tashev et al (20190318755) teaches reduction of feature size while maintaining accuracy:
because of local similarity of a spectrogram in adjacent frequency bins, when convolving with the kernel z, a stride of size b/2 may be used along the frequency dimension. As discussed in detail below, such a design may reduce a number of parameters and computation needed in the recurrent component of the convolutional-recurrent neural network without losing any prediction accuracy.
Yang et al (20240185830) teaches the well know technique of modifying/altering the feature set of the reference speakers, for computational savings:
[0042] where A∈[AltContent: rect].sup.N.sup.p.sup.×N.sup.p denotes a sub-graph representing the feature of the target speaker (for example, the ith speaker as described above), and N.sub.p and N.sub.n denote dimension sizes of features of the target speaker and another speaker (for example, the jth speaker as described above).
Carmiel et al (20230352001) teaches the computation of intermediate acoustic representations altering snippet length, pronunciation, and phonetics – para 0146 – 0152.
However, none of these representative prior art documents, nor other prior art of record, explicitly teach the claim features toward, the smaller target data set relative to the reference data set, of speech features, calculating a third data set which is a subset of the second data set, training a reference set based on the subset third data set, training the acoustic model with the target set and using that output to trains the acoustic model, in conjunction with the generative model using the target data set and the modified/trained reference-based data sets.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Please see related art listed on the PTO-892 form.
Furthermore, see the detailed discussion of representative references, above, in the Reasons for Indicating Allowable Subject Matter.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Michael Opsasnick, telephone number (571)272-7623, who is available Monday-Friday, 9am-5pm.
If attempts to reach the examiner by telephone are unsuccessful, the examiner's supervisor, Mr. Richemond Dorvil, can be reached at (571)272-7602. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
/Michael N Opsasnick/Primary Examiner, Art Unit 2658 6/4/2026
Prosecution Insights