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 .
Response to Arguments
Applicant’s arguments filed on February 02, 2026 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Response to Amendment
The amendment to the claims received on February 02, 2026 has been entered.
The amendment of claims 1, 10 and 19 is acknowledged.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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 1-6, 10-15 and 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Huo’249 (US 2013/0251249), and further in view of Raymond’809 (GB 2573809).
With respect to claim 10, Huo’249 teaches a handwriting recognition apparatus [regarding to the system shown in Fig.1), comprising:
a memory (Fig.2, item 206) operable to store computer-readable instructions (paragraph 34); and
a processor circuitry (Fig. 2, item 202) operable to read the computer-readable instructions, the processor circuitry when executing the computer-readable instructions is configured to (paragraph 30):
obtain handwritten original trajectory data in real-time (paragraphs 34 and 80);
compress the handwritten original trajectory data, to obtain compressed handwritten trajectory data [the number of dimensions of the extracted features of each training character is being reduced by a dimension reduction module to obtain lower-dimensional features (paragraph 45)]; and
input the compressed handwritten trajectory data into a compressed handwriting recognition model for recognition [The recognition model is being constructed according to the generated lower-dimensional features (paragraph 79 and Fig.4, step 408)], to obtain a text recognition result corresponding to the handwritten original trajectory data, a handwriting recognition model a trained model being obtained by training with handwritten trajectory data of each piece of training data in a training data set [the trained character recognition system is being used to obtain text recognition result corresponding to the handwritten original trajectory data (Fig.5, steps 502, 504 and 506, and paragraphs 80-82)], the handwriting recognition model being an end-to-end model inputted with compressed handwritten trajectory data as an input and outputting a text recognition result [the character recognition system (Fig.2, item 102) is considered as the end-to-end model since it being trained and constructed according to the received lower-dimensional features (Fig.4) and it being used to recognize the handwriting trajectory data and then to outputting a text recognition result (Fig.5)].
Huo’249 does not teach the compressed handwriting recognition model being obtained by performing model compression distillation on the handwriting recognition model, a size of the compressed handwriting recognition model being smaller than a size of the handwriting recognition model.
Raymond’809 teaches a first speaker recognition model is compressed using a knowledge distillation compression technique to generate a second speaker recognition model. (page 30). In addition, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to recognize that the size of the second speaker recognition model is considered less than the size of the first speaker recognition model when the said first speaker recognition model is compressed using a knowledge distillation compression technique to generate the said second speaker recognition model since compressing the first speaker recognition model is to reduce its size.
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Huo’249 according to the teaching of Raymond’809 to compress a handwriting recognition model using a knowledge distillation compression technique associate with the generated lower-dimensional features to generate a compressed handwriting recognition model which has smaller size than the said handwriting recognition model for character recognition because this will allow a desired type of handwriting recognition model to be generate and provided for recognize the handwritten more effectively.
With respect to claim 11, which further limits claim 10, Huo’249 teaches wherein the processor circuitry is configured to: perform data preprocessing on handwritten input data that is obtained in real- time, the data preprocessing comprising re-sampling (paragraph 35); and
obtain the handwritten original trajectory data in real-time according to the preprocessed handwritten input data [as shown in Fig. 5, the trained character recognition system with the preprocessed handwritten input data (Fig.2, item 102) obtains the handwritten original trajectory data in real-time (paragraph 35).
With respect to claim 12, which further limits claim 11, Huo’249 teaches wherein the processor circuitry is configured to: perform dimensional compression on the handwritten original trajectory data, to obtain the compressed handwritten trajectory data, a correlation between data of each dimension in the compressed handwritten trajectory data and a model recognition result of the handwriting recognition model being not lower than a predetermined threshold (paragraph 45).
With respect to claim 13, which further limits claim 10, Huo’249 teaches wherein the handwriting recognition model is an end-to-end model [the character recognition system (Fig.2, item 102) is considered as the end-to-end model since it being trained and constructed according to the received lower-dimensional features (Fig.4) and it being used to recognize the handwriting trajectory data and then to outputting a text recognition result (Fig.5].
With respect to claim 14, which further limits claim 13, Huo’249 teaches the processor circuitry is configured to: obtain the training data set and a pre-selected training model corresponding to the training data set (paragraphs 46-48);
obtain the handwritten trajectory data for each piece of training data in the training data set (paragraphs 46-48); and
train the pre-selected training model with the handwritten trajectory data for each piece of training data, to obtain the pre-selected training model that has been trained as the handwriting recognition model (paragraphs 46-48).
With respect to claim 15, which further limits claim 14, Huo’249 teaches wherein the processor circuitry is configured to: obtain a historical handwritten trajectory data set, the historical handwritten trajectory data set comprising at least one of horizontal handwritten trajectory data, vertical handwritten trajectory data, overlapping handwritten trajectory data, and rotating handwritten trajectory data (paragraph 14); and
perform data augmentation on handwritten data in the historical handwritten trajectory data set, and use the data-augmented historical handwritten trajectory data set as the training data set (paragraph 16).
Claims 1-6 are rejected for the same manner as described in the rejected claims 10-15.
Claims 19-20 are rejected for the same manner as described in the rejected claims 10-11.
Claims 7, 8, 16 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Huo’249 (US 2013/0251249), Raymond’809 (GB 2573809) and further in view of Xu’577 (US 2004/0148577).
With respect to claim 16, which further limits claim 15, the combination of Huo’249 and Raymond’809 does not teach wherein the processor circuitry is configured to: obtain a difficult sample and an easy sample in each piece of training data; and train the pre-selected training model in a mode of first training the difficult sample and then training the easy sample.
Xu’577 teaches wherein the processor circuitry is configured to: obtain a difficult sample and an easy sample in each piece of training data [As shown in Fig.1, the handwriting samples are being inputted to the system to perform training and the handwriting samples contain a number of handwritten characters, such as letters, numbers, punctuation marks, as so forth (paragraph 46). Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to recognize to collect the handwriting samples including difficult samples and easy sample as needed to train the system to perform handwriting recognition because this will allow the handwriting recognition model to be trained more effectively to more recognize handwriting more accurately]; and
train the pre-selected training model in a mode of first training the difficult sample and then training the easy sample [As shown in Fig.1, the handwriting samples are being inputted to the system to perform training and the handwriting samples contain a number of handwritten characters, such as letters, numbers, punctuation marks, as so forth (paragraph 46). Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to recognize to collect the handwriting samples including difficult samples and easy sample as needed to train the system either training with the difficult sample first and with the easy sample or training with the easy sample first and with the difficult sample because this will allow the handwriting recognition model to be trained more effectively to more recognize handwriting more accurately].
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Huo’249 and Raymond’809 according to the teaching of Xu’577 to train the handwriting recognition model according to the handwriting samples because this will allow the text associated with handwritten to be recognized more effectively.
With respect to claim 17, which further limits claim 16, Huo’249 teaches use the pre-selected training model that has been trained as the handwriting recognition model [a recognition model is being trained according to the pre-process module (paragraph 35). Therefore, a training model is considered being pre-selected].
The combination of Huo’249 and Raymond’809 does not teach wherein the processor circuitry is configured to: fine-tune the pre-selected training model during a process of training the pre-selected model.
Xu’577 teaches wherein the processor circuitry is configured to: fine-tune the pre-selected training model during a process of training the pre-selected model [the user makes manual adjustment (paragraph 110). Therefore, when the user makes manual adjustment, the training model including any pre-selected training model is considered being updated]
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Huo’249 and Raymond’809 according to the teaching of Xu’577 to train the handwriting recognition model according to the handwriting samples because this will allow the text associated with handwritten to be recognized more effectively.
Claims 7-8 are rejected for the same manner as described in the rejected claims 16-17.
Claims 9 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Huo’249 (US 2013/0251249), Raymond’809 (GB 2573809), Xu’577 (US 2004/0148577) and further in view of Du’167 (CN 111738167).
With respect to claim 18, which further limits claim 16, the combination of Huo’249, Raymond’809 and Xu’577 does not teach wherein the processor circuitry is further configured to: perform model distillation on the pre-selected training model that has been trained, to obtain the distillated pre-selected training model, and use the distillated pre-selected training model as the compressed handwriting recognition model.
Du’167 teaches wherein the processor circuitry is further configured to: perform model distillation on the pre-selected training model that has been trained, to obtain the distillated pre-selected training model [regarding to the distillation gating recursive neural network (paragraph 7)], and
use the distillated pre-selected training model as the compressed handwriting recognition model [a multi-layer distillation GRU network is being used to extract text characteristic (abstract) to perform handwriting recognition (abstract)].
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Huo’249, Raymond’809 and Xu’577 according to the teaching of Du’167 to distillate the handwriting recognition model and then to used it to perform handwriting recognition because this will allow the text associated with handwritten to be recognized more effectively.
Claim 9 is rejected for the same manner as described in the rejected claim 18.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 extension fee 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 date of this final action.
Contact
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HUO LONG CHEN whose telephone number is (571)270-3759. The examiner can normally be reached on M-F 9am - 5pm. If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Tieu, Benny can be reached on (571) 272-7490. 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). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/HUO LONG CHEN/Primary Examiner, Art Unit 2682