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 .
Informality
In claim 14, line 10, delete “ the number of ”, in the first occurrence since the phrase appears in duplicate.
Claims 4, 8, 11, 15 and 22 are cancelled.
Claim Rejections - 35 USC § 101
Claims 17 and 25 are rejected under 35 U.S.C. 101 because the claims are directed to non-patentable subject matter.
At page 34, second full paragraph, of the instant specification, it recites: “… a readable medium may be a readable signal medium or a readable storage medium.” At the top of page 35, the specification states that “… software may be distributed in a computer-readable medium, and the computer readable medium may include a computer storage medium )or a non-transitory medium) and a communication medium (or a transitory medium).
This rejection is supported under MPEP 2106.03.
Non-limiting examples of claims that are not directed to any of the statutory categories include:
Products that do not have a physical or tangible form, such as information (often referred to as "data per se") or a computer program per se (often referred to as "software per se") when claimed as a product without any structural recitations;
Transitory forms of signal transmission (often referred to as "signals per se"), such as a propagating electrical or electromagnetic signal or carrier wave;
Therefore, based upon the admission by the specification which includes at least pages 34 and 35, the Broadest Reasonable Interpretation encompasses transitory forms of signal transmission and solicits a rejection under § 101. Moreover, a claim to a computer readable medium that can be a compact disc or a carrier wave covers a non-statutory embodiment and therefore solicits a rejection as being directed to non-statutory subject matter. See, e.g., Mentor Graphics v. EVE-USA, Inc., 851 F.3d at 1294-95, 112 USPQ2d at 1134 (claims to a "machine-readable medium" were non-statutory, because their scope encompassed both statutory random-access memory and non-statutory carrier waves).
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) 1 is/are rejected under 35 U.S.C. 103 as being unpatentable over CN109858488 in view of CN112633429.
With respect to claim 1, the 488 reference, at Step 4, teaches a training model, for training a handwriting sample recognition model by suing the training sample of a recognition model, see page 4/15, beginning at line 12 of the Disclosure of the Invention. The 488 reference teaches determining an input text image ( the number 8, or example, in figure 1) for recognizing the trace of written text. The 488 reference teaches that at the bottom of page 5/15, the last five lines, the recognition application module detects the position of handwritten characters as claimed.
The 488 reference teaches a feature extraction layer as set forth at page 9/15. The 488 reference teaches a fast RCNN algorithm performing a two part process or a one step process. However, with both methods, a feature extraction layer is utilized in that feature extraction and classification are performed. See the second full paragraph at page 9/15.
The Examiner contends that the 488 reference also performs a fully connected layer in that the to adjust the number of feature maps to accommodate the hand recognition model. In the first para. of page 9/15, the fast RCNN algorithm performs feature extraction on the original image by using a resource to form a “feature-map”. The channel number corresponds to the different classifications of features in the map. Hence, the fully connected layer is believed to be performed by the RCNN algorithm. second full paragraph at page 9/15, a “classification” step is performed.
The claim refers to a Softmax-layer that obtains the prediction probability values of the written text.
In the first para. at page 9, the classification error loss is minimized using bounding box regression. This appears to be related to the prediction of probability values. For example, in the fifth line of the first paragraph, it states that a candidate region network is used to make predictions which are compared with true values to define probability values. In the second full paragraph of the same page, the 488 reference teaches a Soft-max layer as performed by YOLO family of algorithms wherein the multi-class loss function soft-max cross entropy loss is utilized.
The 488 reference teaches multi-neighborhood merging (by means a circulation network) in the last 12 lines of page 9/15.
The 488 reference also teaches a transcription part for converting probabilities into final labels (recognition results).
What is not specifically stated by the 488 reference is the spatial position of the width and height of the written text on a pixel basis. However, this is presumed to be the case if the hand written text is represented digitally. Assuming arguendo, the 429 reference is directed to a model for recognizing students hand written choices. The 429 reference also teaches original pictures of the handwritten model is represented as gray scale images normalized to 64 x100. The 429 references states that the gray scale picture is divided by the maximum value of pixels minus .5.
Therefore, the purpose of representing the handwriting samples in terms of pixels would have been recognized by the 488 reference as set forth by the 429 reference.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to represent the spatial relationships of handwritten features in terms
Claim(s) 6 is/are rejected under 35 U.S.C. 103 as being unpatentable over CN109858488 in view of CN112633429 further in view of Fukada (JPH10161997).
The 488 reference in view of the 429 reference teaches all of the subject matter upon which the claim depends, except for determining a number of lines of written text and determining the height of the input text image.
At page 2 of the Fukada reference, it teaches “… determining in each area, the number of lines and digits and the character size …. in advance.” (See para. 3). The Examiner contends that the determining means is performed by the document processing apparatus 10. Therefore, if the total number of lines of text is known and the size of the text is also determined by the processor, then the total height of all lines of text can be determined.
Since, the 488 reference, the 429 reference and the Fukada references are directed to handwriting recognition and processing, the purpose of a measuring or determining the number of lines of text as well as the overall size of the text height, would have been recognized by the 488 and/or the 429 reference.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the optical character recognition module described at the bottom of page 3 of the 429 reference, or to modify the handwriting module, described by the 488 reference, to perform the same functions of measuring or determining the number of lines of text and the height of the text as set forth by Fukada.
Claim(s) 7 is/are rejected under 35 U.S.C. 103 as being unpatentable over CN109858488 in view of CN112633429 further in view of Fukada (JPH10161997) and KSR v. Teleflex 550 U.S. 398 (2007)
With respect to claim 7, CN109858488 in view of CN112633429 further in view of Fukada (JPH10161997) teaches the determination of the number of line of text and the height of the overall text given the computation of the font or character size.
What is not specifically taught are the values of input_h, raw_num, [ ] and other input the claims receipt to define the height of the input text.
However, it would have been recognized by one of ordinary skill in the art to try different methods, using different formulae and computations for the purpose of determining the height of a text based on the number or lines and the size of the fonts.
Claim(s) 12 ,13, 16, 17, 24 and 25 is/are rejected under 35 U.S.C. 103 as being unpatentable over CN109858488 in view of CN112633429 further in view of Lee (2022/0138453).
With respect to claim 12, the 488 reference and the 429 reference teaches all of the subject matter upon which claim 12 depends, as recited in the rejection to claim 1. What is not taught by either reference is correcting English words in accordance with a pre-established corpus.
Lee teaches the use of a character-based language model that improves handwriting recognition, see para. 39. At para 40, Lee using a library of misspelled words to assist in the identification of the correct spelling. For example, see para. 47 where a prediction is made to correctly spell the name of “Warren” after receiving “Warran”. See also para. 86 for correcting English words (US data set).
With respect to claim 13, Lee teaches detecting whether English words are in the corpus or not by use of a by use of a US name dataset, see para. 39, last three lines. Lee also teaches correcting words that are misspelled. For example, see the correction of the spelling to of Warren when the misspelling identified as Warran was made. See para. 47.
Lee also teaches correcting words within a edit distance, which appears to be “L2 distance” as set forth in para. 38, last four lines.
Hence spelling errors may be corrected with handwriting recognition embeddings from the character-based language model, see para. 43. The motivation for this rejection is the same as that to claim 12.
With respect to claim 16, the 488 reference and the 429 reference teaches all of the subject matter upon which claim 16 depends, as recited in the rejection to claim 1. What is not taught by either reference is use of a memory for storing instructions operable by a processor to implement the steps of claim 1.
Lee teaches at last one processor 504 storing instructions (commands 503 or computer programs, see para. 114) for implementing the handwriting recognition module (for example, 605 or 610). Lee teaches handwriting model for correcting spelling errors for providing an automated approach to augmenting, refining and improving the handwriting model to predict English written text.
Since Lee provides the motivation for using a computer, with a processor and storing instructions in a memory to implement the method steps, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use a computer, with a processor and memory for providing instructions to run the Neural networks as set forth in the 488 and the 429 references in the same manner that neural networks are operated with the processor, computer and memory with stored programming, as set forth by Lee.
With respect to claim 17, the 488 reference and the 429 reference teaches all of the subject matter upon which claim 17 depends, as recited in the rejection to claim 1. What is not taught by either reference is use of a computer readable medium for storing a program having instructions for performing the steps of claim 1.
Lee teaches a computer system 500 with one or more processors 504, a computer readable storage medium 501 for storing instructions 503 for implementing handwriting recognition. Lee teaches handwriting model for correcting spelling errors for providing an automated approach to augmenting, refining and improving the handwriting model to predict English written text.
Since Lee provides the motivation for using a computer, with a processor and storing instructions in a memory to implement the method steps, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use a computer readable storage medium for storing instructions that are implemented with a processor to run the Neural networks as set forth in the 488 and the 429 references in the same manner that neural networks are operated with the processor, computer and memory with stored programming, as set forth by Lee.
With respect to claim 24, the 488 reference and the 429 reference teach training but does not show the use of a memory and a processor wherein the memory contains instructions that are implemented by the processor for performing the handwriting module steps.
Lee teaches a computer system 500 with one or more processors 504, a computer readable storage medium 501 for storing instructions 503 for implementing handwriting recognition. Lee teaches handwriting model for correcting spelling errors for providing an automated approach to augmenting, refining and improving the handwriting model to predict English written text.
At paragraphs 38, 39, Lee teaches that the character-based language models can be trained by providing a pair of words to the character based learning model. Lee also teaches that the character-based language model may be trained or adopted using a suitable data set, (para. 39, lines 1-3). Moreover, at para. 123, Lee teaches that the handwriting recognition model 605 can be trained according to the embodiments.
At the bottom of paras. 114 and 121, Lee teaches that the use of the processor(s) 504 and memory 501 with stored instructions 503, may be used to perform one or more steps or any of the methods or embodiments described with the Lee reference.
Since Lee provides the motivation for using a computer, with a processor and storing instructions in a memory to implement the method steps, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use a computer, with a processor and memory for providing instructions to train the handwriting language model. It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the 488 and/or the 429 reference to implement the handwriting language models of these references, with those used by Lee, in the manner described by Lee, to make the claimed invention.
With respect to claim 25, the 488 reference and the 429 reference teach all of the claimed subject matter as set forth in the rejection to claim 1.
However, the combined references do not show the use of a computer readable medium with a computer program executed by the model for performing a training in accordance with claim 1.
Lee teaches a computer program (instructions 503), that are stored within a storage medium (501) that are executed by at least one processor(s) (504). Lee teaches handwriting model for correcting spelling errors for providing an automated approach to augmenting, refining and improving the handwriting model to predict English written text.
At paragraphs 38, 39, Lee teaches that the character-based language models can be trained by providing a pair of words to the character based learning model. Lee also teaches that the character-based language model may be trained or adopted using a suitable data set, (para. 39, lines 1-3). Moreover, at para. 123, Lee teaches that the handwriting recognition model 605 can be trained according to the embodiments.
At the bottom of paras. 114 and 121, Lee teaches that the use of the processor(s) 504 and memory 501 with stored instructions 503, may be used to perform one or more steps or any of the methods or embodiments described with the Lee reference.
Since Lee provides the motivation for using a computer, with a processor and storing instructions in a memory to implement the method steps, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use a computer, with a processor and memory for providing instructions to train the handwriting language model. It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the 488 and/or the 429 reference to implement the handwriting language models of these references, with those used by Lee, in the manner described by Lee, to make the claimed invention.
Claims Objected As Containing Allowable Matter
Claims 2, 3, 5, 9, 10 and 14 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Claim 2 is objected to as containing allowable matter for the reason the prior art does not teach in claimed combination, constructing a training model which is comprised of a handwritten recognition model and a height compression module wherein the height compression module is between the image feature extraction layer and the full connection layer and removing the height compression module to obtain a trained handwriting recognition module.
Claims 3 and 5 are allowed by their direct or indirect relation from claim 2.
Claim 9 is objected to as containing allowable matter for the reason the prior art does not teach in claimed combination, calculating a scaled factor ratio between the input text image and the written text trace and determining coordinates of trace points in the input text image.
Claim 10 is objected to as containing allowable matter for the reason the prior art does not teach or suggest in claimed combination, performing same-line alignment on the prediction results of different spatial positions by means of calculation of the average value of X axis coordinates and average value of Y axis coordinates of all pixels in each connected domain after multi-neighborhood merging and traversing the connected domain and aligning pixels with a difference in average y of the Y axis of less than 5 pixels.
Claim 14 is objected to as containing allowable subject matter for the reason the prior art does not teach or suggest in claimed combination, “…. self-increasing the current detection value of the minimum edit distance by 1 when the first English word does not exist, and returning to the step of detecting whether the first English word exists and the number of the first English word to perform cyclic detection until the current detection value of the minimum edit distance is greater than a predefined threshold value of the minimum edit distance….”
Allowable Subject Matter
Claims 18-21 and 23 allowed.
Claim 18 is allowed for the reason the prior art does not teach or suggest, height compression module is disposed between the image feature extraction layer and the fully connection layer for compressing a height of the feature map extracted by the image feature extraction layer.
Claim 19 is allowed by its dependency from claim 18.
Claim 20 is allowed by its dependency from claim 18.
Claim 21 is allowed by its dependency from claim 20.
Claim 23 is allowed by its dependency from claim 18.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JEROME GRANT II whose telephone number is (571)272-7463. The examiner can normally be reached M-F 9:00 a.m. - 5:00 p.m..
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, Jennifer Mehmood can be reached at 571-272-2976. 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.
/JEROME GRANT II/Primary Examiner, Art Unit 2664