DETAILED ACTION
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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 03/27/2026 has been entered.
Response to Arguments
Applicant's arguments filed 03/27/2026 have been fully considered but they are not fully persuasive.
Regarding the 101 rejections, applicant’s arguments and amendments to the independent claims are persuasive and overcome the previous 101 rejections. Specifically, applicant’s amended limitations receiving user input specifying a directionality of alteration for extending original music data by concatenation of input music features; generating new extension music data by using the user input received with respect to the original music data, a plurality of input music features derived from the original music data that are not in a concatenating relationship, and a trained model that is trained on inputs including a feature sampled from a standard normal distribution of the original music data, wherein the new extension music data is obtained from the plurality of input music features having alterations to selected input music features determined based on the directionality specified by the user input, wherein the alterations facilitate concatenation of musical elements corresponding to the selected input music features of the plurality of input music features with musical continuity between the new extension music data and the original music data provides a technical improvement. Using a model, trained on a standard normal distribution of original music data, to generate extension music, that has musical continuity to the original music data, ensures that the new extension music does not add unnaturalness and discontinuity to the music. See pg. 14 of “Remarks”: “Applicant respectfully submits that the Examiner has also "oversimplified the [technical] component of the claims and downplayed the invention's benefits." See Enfish. Moreover, similar to Desjardins, the presently claimed invention provides a specific technical solution by use of at least a trained model that is trained on inputs including a feature sampled from a standard normal distribution of the original music data, along with other elements recited in relation to the model.” See paragraph 15 of the Specification: “The new data is data obtained from a plurality of features having alterations. The plurality of features with an alteration can be novel features that cannot be obtained by simply concatenating the individual features before the alteration while holding distinct characteristics of the plurality of features before the alteration. This can also reduce unnaturalness such as discontinuity that can occur when the individual features before the alteration are simply concatenated.” Applicant’s amendments and corresponding arguments that the claimed invention provides a technical improvement to the field of music generation are persuasive. Therefore, the 101 rejections are withdrawn.
Regarding the 103 rejections, applicant's arguments filed with respect to the prior art rejections have been fully considered but they are moot. Applicant has amended the claims to recite new combinations of limitations. Applicant's arguments are directed at the amendment. Please see below for new grounds of rejection, necessitated by Amendment.
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.
Claims 1-13 are rejected under 35 U.S.C. 103 as being unpatentable over Kolen, et al., US Patent Publication US10657934B1 (“Kolen”) in view of Simon, et al., “LEARNING A LATENT SPACE OF MULTITRACK MEASURES” (“Simon”) and further in view of Zhang, Non-Patent Literature “A Short Guide for Feature Engineering and Feature Selection” (“Zhang”).
Regarding claim 1, Kolen discloses:
An information processing method for generating music data, the method comprising: receiving user input specifying a directionality of alteration for extending original music data by concatenation of input music features; (Kolen, col. 1 lines 56-62, “As will be described, a musical composition application [An information processing method for generating music data,] may advantageously recommend musical phrases or passages based on previously input musical notes. For example, the application may recommend minutes, or even hours, of a musical score [for extending original music data by concatenation of input music features;]. Additionally, the musical composition application may adjust musical notes specified by a user according to different constraints [the method comprising: receiving user input specifying a directionality of alteration].”).
generating new extension music by using the user input received with respect to the original music data, a plurality of input music features derived from the original music data…, and a trained model that is trained on inputs including a feature sampled from…the original music data, wherein the new extension music data is obtained from the plurality of input music features having alterations to selected input music features determined based on the directionality specified by user input, (Kolen, col. 2-3, “Indeed, and as will be described, a composer may specify a certain theme or melody. The musical composition application may then expound upon this specified theme or melody. For example, the musical score application may generate one or more measures for inclusion in the musical score [generating new extension music by using the user input received with respect to the original music data,]. Advantageously, these generated measures may conform to a same musical style as being utilized by the composer. Thus, the musical score application may rapidly auto-complete a musical score being created by the composer [a plurality of input music features derived from the original music data… wherein the new extension music data is obtained from the plurality of input music features having alterations to selected input music features determined based on the directionality specified by user input,].”, and Kolen, col. 4 lines 38-40, “the recommended musical phrase being determined based on one or more machine learning models [and a trained model]”, and Kolen, col. 7-8, “ the system may learn constraints utilizing one or more machine learning models. For example, an artificial neural network (e.g., a recurrent neural network) may be utilized to learn common sequences of features. In this example, the system may train the artificial neural network on different aspects of the artist's songs. For example, system may encode certain features, such as chord changes, utilization of leading tones, ands son, as training data. The system may then train an artificial neural network to recognize these features [that is trained on inputs including a feature sampled from…the original music data,].”).
wherein the alterations facilitate concatenation of musical elements corresponding to the selected input music features of the plurality of input music features with musical continuity between the new extension music data and the original music data, (Kolen, col. 2-3, “Indeed, and as will be described, a composer may specify a certain theme or melody. The musical composition application may then expound upon this specified theme or melody. For example, the musical score application may generate one or more measures for inclusion in the musical score. Advantageously, these generated measures may conform to a same musical style as being utilized by the composer. Thus, the musical score application may rapidly auto-complete a musical score being created by the composer [wherein the alterations facilitate concatenation of musical elements corresponding to the selected input music features of the plurality of input music features with musical continuity between the new extension music data and the original music data,].”).
and wherein, upon receiving the user input with respect to the original music data, the trained model is configured to output a plurality of output music features having the alterations; (Kolen, col. 6 lines 49-56, “An example recommendation may include recommended musical notes, for example to complete or expound upon musical notes specified by a user [and wherein, upon receiving the user input with respect to the original music data,]. For example, the system may recommend a particular musical phrase be included in the musical score. The musical phrase may represent a measure of music or may represent minutes or hours of music (e.g., generated via machine learning models) [the trained model is configured to output a plurality of output music features having the alterations;].”).
and displaying the generated new extension music data as an extension of the original music data. (Kolen, col. 4 lines 36-38, “The user interface: presents a recommended musical phrase for inclusion in the musical score [and displaying the generated new extension music data as an extension of the original music data.]”).
While Kolen teaches a music generation program that takes multiple user preferences to extend or modify a musical track, Kolen does not explicitly teach:
that are not in a concatenating relationship
a feature sampled from a standard normal distribution
Simon teaches that are not in a concatenating relationship (Simon, pg. 2 col. 2, “We model measures with up to 8 tracks (see Figure 1 for an example with 4 tracks). Each track consists of a single “instrument” as extracted by pretty midi[31]. A track is represented as a MIDI-like sequence of events from an extension of the vocabulary used by Simon and Oore [37] to handle metric timing and choice of instrument; using separate tracks for each instrument is interpreted as features not in a concatenating relationship as each track correlates to a different instrument and thus not in a concatenating relationship (i.e. that are not in a concatenating relationship)”).
Kolen and Simon are both in the same field of endeavor (i.e. music generation). It would have been obvious for a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Kolen and Simon to teach the above limitation(s). The motivation for doing so is that considering multiple instruments at once improves music creation by considering the harmony between the instruments (cf. Simon, pg. 2 col. 1, “Like our work, the system uses a latent space shared across tracks to handle interdependencies between instruments.”).
While Kolen in view of Simon teaches music generation using altered features from original music data, the combination does not explicitly teach:
a feature sampled from a standard normal distribution
Zhang teaches a feature sampled from a standard normal distribution (Zhang, pg. 13 see Table 3.1.2, “Method: Normalization – Standardization (Z-score scaling)…Definition: removes the mean and scales the data to unit variance. z = (X - X.mean)/ std…Pros: feature is rescaled to have a standard normal distribution that centered around 0 with SD of 1 [a feature sampled from a standard normal distribution]”).
Kolen, in view of Simon and Zhang are both in the same field of endeavor (i.e. machine learning). It would have been obvious for a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Kolen, in view of Simon, and Zhang to teach the above limitation(s). The motivation for doing so is that using a standard normal distribution feature scaling improves training of the model (cf. Zhang, pg. 13, “If range of inputs varies, in some algorithms, object functions will not work properly. Gradient descent converges much faster with feature scaling done. Gradient descent is a common optimization algorithm used in logistic regression, SVMs, neural networks etc.”).
Regarding claim 2, Kolen in view of Simon and Zhang teaches the information processing method according to claim 1. Kolen further teaches wherein the plurality of input music features includes features extracted from partial data having a data length shorter than a data length of the new extension music data. (Kolen, col. 13-14, “The musical phrase, as described herein, may represent a portion of music. For example, a particular measure of music may be generated. As another example, multiple measures may be generated. As another example, a new melody line may be generated. In this example, the user may confirm the new melody line and one or more musical phrases or portions [wherein the plurality of input music features includes features extracted from partial data having a data length shorter] may be generated based on the melody line [than a data length of the new extension music data.].”).
Regarding claim 3, Kolen in view of Simon and Zhang teaches the information processing method according to claim 1. Kolen further teaches:
wherein each of the plurality of input music features is a feature extracted from partial data having a data length shorter than a data length of the new extension music data, (Kolen, col. 13-14, “The musical phrase, as described herein, may represent a portion of music. For example, a particular measure of music may be generated. As another example, multiple measures may be generated. As another example, a new melody line may be generated. In this example, the user may confirm the new melody line and one or more musical phrases or portions [wherein each of the plurality of input music features is a feature extracted from partial data having a data length shorter] may be generated based on the melody line [than a data length of the new extension music data.].”).
and wherein the new extension music data has the same data length as a total data length of each piece of partial data corresponding to each of the plurality of input music features. (Kolen, col. 13-14, “The musical phrase, as described herein, may represent a portion of music. For example, a particular measure of music may be generated. As another example, multiple measures may be generated. As another example, a new melody line may be generated. In this example, the user may confirm the new melody line and one or more musical phrases or portions may be generated based on the melody line; the melody line, or new extension music, is made up of the one or more generated musical phrases or portions (i.e. and wherein the new extension music data has the same data length as a total data length of each piece of partial data corresponding to each of the plurality of input music features.).”).
Regarding claim 4, Kolen in view of Simon and Zhang teaches the information processing method according to claim 1. Simon further teaches:
further comprising: generating additional new extension music data obtained from the plurality of output music features having further alterations, (Simon, pg. 1, col. 2 and see Figure 3, “Apply attribute transformations to an existing measure, e.g. “increase note density” or “add strings” [further comprising: generating additional new extension music data obtained from the plurality of output music features having further alterations,].”).
the generation of the additional new extension music data performed using the plurality of output music features and iteratively using the trained model. (Simon, pg. 6 col. 2, “We have shown how to train and apply a latent space model over measures of symbolic music with multiple polyphonic instruments [performed using the plurality of output music features and iteratively using the trained model.]. We believe that ours is the first model capable of generating full multitrack polyphonic sequences [the generation of the additional new extension music data] with arbitrary instrumentation.”).
Kolen, in view of Zhang, and Simon are both in the same field of endeavor (i.e. music generation). It would have been obvious for a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Kolen, in view of Zhang, and Simon to teach the above limitation(s). The motivation for doing so is that modifications to measures allows for the ability to control and generate music with rich instrumentation (cf. Simon, abstract, “We demonstrate that our latent space model makes it possible to intuitively control and generate musical sequences with rich instrumentation”).
Regarding claim 5, Kolen in view of Simon and Zhang teaches the information processing method according to claim 4. Simon further teaches:
further comprising: displaying the new extension music data and the additional new extension music data that have been generated (Simon, pg. 4 see Figure 2 below,
PNG
media_image1.png
295
887
media_image1.png
Greyscale
Figure 2 displays two measures generated by the model and is interpreted as a new extension and the additional new extension (i.e. further comprising: displaying the new extension music data and the additional new extension music data that have been generated)).
and a number of times of alterations to generate each of the plurality of output music features by the iterative use of the trained model, in association with each other. (Simon, pg. 4 see Figure 3 below,
PNG
media_image2.png
287
912
media_image2.png
Greyscale
“Figure 3. Multiple transformations to a single measure via attribute vector arithmetic. On the left is the original measure, followed by its reconstruction from the latent space. After that are three transformations: increasing the pitch range, using only string instruments, and using more tracks [and a number of times of alterations to generate each of the plurality of output music features by the iterative use of the trained model, in association with each other.]”).
Kolen, in view of Zhang, and Simon are all in the same field of endeavor (i.e. music generation). It would have been obvious for a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Kolen, in view of Zhang, and Simon to teach the above limitation(s). The motivation for doing so is that displaying the results of a combination or change can inform the user about the progress of a music track.
Regarding claim 6, Kolen in view of Simon and Zhang teaches the information processing method according to claim 1. Kolen further teaches further comprising: generating the new extension music data by also using an additional feature determined with respect to the plurality of input music features. (Kolen, col. 8 lines 10-15, “As the user utilizes the music composition application, the system may recommend one or more musical phrases for inclusion in a musical score [further comprising: generating the new extension music data]. The recommended musical phrases may be based, at least in part, on the musical genre. For example, the system may utilize the determined constraints to generate a recommended musical phrase [by also using an additional feature determined with respect to the plurality of input music features.].”).
Regarding claim 7, Kolen in view of Simon and Zhang teaches the information processing method according to claim 6. Kolen further teaches further comprising: displaying the new extension music data that has been generated and the directionality of an alteration given by the additional feature, in association with each other. (Kolen, see Figure 1D, Figure 1D shows the recommended generated musical phrase based on the genre and constraints given by the user, therefore the new extension music is displayed with the directionality of an alteration given by the additional feature, the genre (i.e. further comprising: displaying the new extension music data that has been generated and the directionality of an alteration given by the additional feature, in association with each other.)).
Regarding claim 8, Kolen in view of Simon and Zhang teaches the information processing method according to claim 6. Kolen further teaches further comprising: displaying the additional feature corresponding to the new extension music data that has been generated. (Kolen, see Figure 1D, Figure 1D shows the recommended generated musical phrase based on the genre and constraints given by the user, therefore the additional feature, the genre, is displayed with the new extension music (i.e. further comprising: displaying the additional feature corresponding to the new extension music data that has been generated.)).
Regarding claim 9, Kolen in view of Simon and Zhang teaches the information processing method according to claim 1. The combination also teaches the original music data as seen in claim 1. Zhang further teaches wherein the plurality of input music features includes one or more features sampled from the standard normal distribution of the original music data. (Zhang, pg. 13 see Table 3.1.2, “Method: Normalization – Standardization (Z-score scaling)…Definition: removes the mean and scales the data to unit variance. z = (X - X.mean)/ std…Pros: feature is rescaled to have a standard normal distribution that centered around 0 with SD of 1 [wherein the plurality of input music features includes one or more features sampled from the standard normal distribution of the original music data.]”).
It would have been obvious to one of ordinary skill in the art before the effective filling date of the present application to combine the teachings of Zhang with the teachings of Kolen and Simon for the same reasons disclosed in claim 1.
Regarding claim 10, Kolen in view of Simon and Zhang teaches the information processing method according to claim 9. The combination also teaches the original music data as seen in claim 9. Kolen further teaches a feature extracted from partial data having a data length shorter than a data length of the new extension music data. (Kolen, col. 13-14, “The musical phrase, as described herein, may represent a portion of music. For example, a particular measure of music may be generated. As another example, multiple measures may be generated. As another example, a new melody line may be generated. In this example, the user may confirm the new melody line and one or more musical phrases or portions [a feature extracted from partial data having a data length shorter] may be generated based on the melody line [than a data length of the new extension music data.].”).
Zhang further teaches wherein a feature sampled from the standard normal distribution of the original music data is used instead of (Zhang, pg. 13 see Table 3.1.2, “Method: Normalization – Standardization (Z-score scaling)…Definition: removes the mean and scales the data to unit variance. z = (X - X.mean)/ std…Pros: feature is rescaled to have a standard normal distribution that centered around 0 with SD of 1 [wherein a feature sampled from the standard normal distribution of the original music data]” and Zhang, pg. 13, “If range of inputs varies, in some algorithms, object functions will not work properly. Gradient descent converges much faster with feature scaling done [is used instead of]. Gradient descent is a common optimization algorithm used in logistic regression, SVMs, neural networks etc.”).
It would have been obvious to one of ordinary skill in the art before the effective filling date of the present application to combine the teachings of Zhang with the teachings of Kolen and Simon for the same reasons disclosed in claim 9.
Regarding claim 11, the claim is similar to claim 1. Kolen teaches the additional limitations An information processing apparatus comprising: circuitry configured to (Kolen, col. 15 lines 6-9, “The musical adjustment and recommendation system 200 may be a system of one or more computers, one or more virtual machines executing on a system of one or more computers, and so on [An information processing apparatus comprising: circuitry configured to].”).
Regarding claim 12, the claim is similar to claim 1. Kolen teaches the additional limitations A non-transitory computer-readable storage medium having embodied thereon an information processing program, which when executed by a computer causes the computer to function execute a method for generating music data (Kolen, col. 4 lines 52-54, “Some aspects feature a computing system comprising one or more processors and non-transitory computer storage media storing instructions that when executed by the one or more processors [A non-transitory computer-readable storage medium having embodied thereon an information processing program, which when executed by a computer causes the computer to function execute a method for generating music data]”).
Regarding claim 13, Kolen in view of Simon and Zhang teaches the information processing method according to claim 1. Kolen further teaches wherein the generated new extension music data is displayed in association with the directionality of alteration specified by the user input. (Kolen, col. lines and see Figure 1D, “FIG. 1D illustrates the user interface 100 presenting a recommended musical phrase for inclusion in the musical score 106. The user interface 100 may optionally present musical phrases for inclusion in the musical score 106. For example, the musical adjustment and recommendation system 200 may utilize machine learning techniques to generate a musical phrase; Figure 1D shows that the directionality of alteration provided by the user in element 104 is visible with the generated new extension music in element 144 (i.e. wherein the generated new extension music data is displayed in association with the directionality of alteration specified by the user input.).”).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Sawruk, et al., US20180341702A1 discloses a system that searches or discovers musical sheet music using user input constraints. The system leverages the metadata within different musical compositions to generate sheet music based on the user’s constraints.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to NICHOLAS S WU whose telephone number is (571)270-0939. The examiner can normally be reached Monday - Friday 8:00 am - 4:00 pm EST.
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, Michelle Bechtold can be reached at 571-431-0762. 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.
/N.S.W./Examiner, Art Unit 2148
/MICHELLE T BECHTOLD/Supervisory Patent Examiner, Art Unit 2148
Prosecution Insights