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 .
Priority
Receipt is acknowledged of certified copies of papers submitted under 35 U.S.C. 119(a)- (d), which papers have been placed of record in the file.
Specification
The specification has not been reviewed in detail to determine the presence of all possible minor errors. Applicant's attention is directed to the specification, and applicant is requested to make any necessary corrections of which applicant is aware.
Status of Claims
This Office Action is in response to the application filed on 05/29/2026. Claim 1-36 are cancelled. Claims 37-58 are pending and have been examined.
Information Disclosure Statement
The information disclosure statements (IDSs) submitted on 11/19 /2024,11/25/2024, 05/30/2025, and 08/14/2025 have been filed in accordance with the provisions of 37 CFR 1.97. Accordingly, they are being considered by the examiner.
Claim interpretation
Claim 55 recites a non-transitory computer-readable recording medium comprising a bitstream obtained by using an encoding method.
The phrase “obtained by using an encoding method” describes the manner in which the bitstream is produced and is interpreted as product-by-process language, where the product is the bitstream and the recited encoding method is the process used to obtain the bitstream.
MPEP §2113 explains that “Product-by-Process claims are not limited to the manipulations of the recited steps, but rather to the structure implied by those steps”. Thus, the scope of the claim is directed to the non-transitory computer-readable storage medium comprising the resulting bitstream rather than to the performance of the encoding method itself.
Accordingly, claim 55 is interpreted as being directed to a non-transitory computer readable storage medium comprising a bitstream obtained by an encoding method.
2111.05 Functional and Nonfunctional Descriptive Material [R-07.2022]
The claimed bitstream does not impart functionality to a computer or computing device, and is therefore considered nonfunctional descriptive material. Such nonfunctional descriptive material, in the absence of a functional relationship with a computer-readable medium, does not constitute a statutory process, machine, manufacture or composition of matter.
A computer-readable storage medium (CRM) and the bitstream (i.e. descriptive material) must be in a functional relationship to be given a patentable weight. A functional relationship exists where the descriptive material performs a function with respect to the medium. See MPEP §2111.05(I)(A).
When a computer-readable medium merely serves as a support for information or data, no functional relationship exists. See MPEP §2111.05(II).
Here, the computer-readable medium merely storing the bitstream and provides no functional relationship between the stored bitstream and medium. Accordingly, the claimed subject matter does not fall within one of the four statutory categories of invention and therefore non-statutory.
Claim Rejections - 35 USC § 102
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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claim 55 is rejected under 35 U.S.C. 102(a)(1) as being anticipated by over Kim (US- 11483476-B2) hereinafter “Kim”.
Regarding Claim 55 Kim
Kim discloses 55. (New) A non-transitory computer-readable storage medium comprising a bitstream obtained by using an encoding method ( Kim, Col. 116, lines18-20 “A non-transitory computer-readable storage medium storing a bitstream generated by a method of encoding an image, the method ….” comprising:
The remaining limitations directed to the encoding method are not given patentable weight, as explained in the §101 rejection and the claim interpretation section above.
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.
The factual inquiries for establishing a background for determining obviousness under 35
U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness
Claims 37, and 53-58 are rejected under 35 U.S.C. 103 as being unpatentable over Zhou et al (US 20200374522 A1) hereinafter “Zhou” in view of Flynn et al (US-20210105022-A1) hereinafter “Flynn”.
Regarding Claim 37 Zhou-Flynn
Zhou discloses 1.(New) An encoding method, (Fig. 9; [0157] “an image coding method. FIG. 9 is a schematic diagram of a process of the image coding method …”) comprising:
obtaining to-be-encoded feature data comprising a plurality of feature elements of a picture feature map or an audio feature variable; (Zhou, [0158] 901: feature extraction is performed on to-be-processed image data by using a convolutional neural network, to generate feature maps of the image data; ”) and
for a first feature element in the plurality of feature elements: ( Zhou, [0089] “..part of the data may be selected as the preprocessed data from the discrete feature maps. .. the preprocessed data include a plurality of data within a specified range centered on the current data to be coded.” [0124] As shown in FIG. 5, first, feature extraction is performed on the to-be-processed image data…”)
determining whether entropy encoding needs to be performed on the first feature element; (Zhou, [0196] “…in performing entropy coding on the data to be coded in a channel in the discrete feature maps according to the flag data….when the flag data indicate that the data of a channel in the discrete feature maps are all zero, no entropy coding is performed on the data to be coded of the channel in the discrete feature maps; and when the flag data indicate that data of a channel in the feature maps are not all zero, entropy coding is performed on the data to be coded according to the probabilities of the data to be coded of the channel.”)
Zhou does not expressly disclose
when the entropy encoding needs to be performed on the first feature element, performing the entropy encoding on the first feature element and writing a result of the entropy encoding performed on the first feature element into a bitstream after performing the entropy encoding; and
when the entropy encoding does not need to be performed on the first feature element, skipping performing the entropy encoding on the first feature element and writing data into the bitstream for the first feature element.
However, in the same field of endeavor Flynn discloses more explicitly the following:
when the entropy encoding needs to be performed on the first feature element, performing the entropy encoding on the first feature element and writing a result of the entropy encoding performed on the first feature element into a bitstream after performing the entropy encoding; (Flynn, [0040] “encoder 100, may output a sequence of packets to represent bitstream 170” [0042] “…encoded symbols from a compressible symbol sub-stream 122 that have been entropy encoded by entropy encoder 108 may be written directly to a packet, such as packet 130. For example, the entropy encoded symbols may be written in a forward order in entropy encoded bytes 136.”) and
when the entropy encoding does not need to be performed on the first feature element, skipping performing the entropy encoding on the first feature element and writing data into the bitstream for the first feature element. (Flynn, [0022] “…a sub-stream of symbols to be entropy encoded by an entropy encoding component of an encoder, and instead directing the non-compressible symbols into a bypass sub-stream.” [0042] “bypass symbols from non-compressible symbol sub-stream 124 may be written directly to packet 130 in a reverse order in bypass bytes 138.” since the packet represents the bitstream, this teaches writing bypass data into the bitstream when entropy encoding is skipped.)
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the application to modify the teachings of Zhou in view of Flynn to write entropy-encoded data into the bitstream when entropy is performed, and to write bypass data into the bitstream when entropy encoded is skipped, because Flynn teaches that entropy encoding may be a bottleneck and that symbols that do not benefit from entropy encoded may be directed into a bypass sub-stream to reduce latency and computational cost while still forming a packet for transmission or storage. (Flynn, ¶¶ [0021]-[0023])
Note: The motivation that was utilized in the rejection of claim 1, applies equally as well to claims 53-58.
Regarding Claim 53 Zhou-Flynn
The independent claim 53 recites a limitation that are substantially the similar to those of independent claim 37, except that claim 53 is directed to decoder rather than an encoder. It is well established in the art that video compression systems comprise complementary components, namely encoder (compressor) and a decoder (decompressor), which perform reciprocal operations. The encoder compresses source data to reduce the bit rate for transmission or storage, while the decoder reconstructs the data from the compressed bitstream by performing a corresponding inverse process.
Regarding Claim 54 Zhou-Flynn
Zhou discloses 54. (New) A computer program product comprising program code, wherein when the program code is executed by a computer or a processor, the program code causes the computer or the processor to: (Zhou, [0250] “An embodiment of the present disclosure provides a computer storage medium, including a computer readable program code, which will cause a computer to carry out the image coding method described in the first aspect or the third aspect of the embodiments in an image coding apparatus.”)
The remaining limitations of independent claim 54 recite features that are substantially similar to those set forth in independent claim 37. Accordingly, the reasoning and analysis provided with respect to claim 37 apply equally to claim 54.
Regarding Claim 56 Zhou-Flynn
Zhou-Flynn discloses 56. (New) An encoder, (Zhou [0033] “FIG. 1B illustrates an example chunk or packet encoded by the encoder…”) comprising:
a non-transitory computer-readable storage medium storing a program; and one or more processors coupled to the non-transitory computer-readable storage medium and configured to execute the program to cause the encoder to: (Flynn, Claim 1, “A non-transitory computer-readable medium storing program instructions that, when executed by the one or more processors, cause the one or more processors to: separate a stream of symbols to be entropy encoded into….”(Zhou, [0240] “FIG. 15 is a schematic diagram of a structure of the electronic device of the embodiment of this disclosure. As shown in FIG. 15, an electronic device 1500 may include a processor (such as a central processing unit (CPU)) 1510 and a memory 1520, the memory 1520 being coupled to the central processing unit 1510. The memory 1520 may store various data, and furthermore, it may store a program for information processing, and execute the program under control of the processor 1510.”)
The remaining limitations of independent claim 56 recite features that are substantially similar to those set forth in independent claim 37. Accordingly, the reasoning and analysis provided with respect to claim 37 apply equally to claim 56.
Regarding Claim 57 Zhou-Flynn
Zhou-Flynn discloses 57. (New) A picture or audio processor comprising a processing circuit and configured to: (Zhou, [0060] “FIG. 5 is a schematic diagram of a structure of the image coding apparatus.” [0255] “One or more functional blocks and/or one or more combinations of the functional blocks in the drawings may be realized as a universal processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices,…”)
The remaining limitations of independent claim 57 recite features that are substantially similar to those set forth in independent claim 37. Accordingly, the reasoning and analysis provided with respect to claim 37 apply equally to claim 57.
Regarding Claim 58 Zhou-Flynn
Zhou-Flynn discloses 58. (New) A non-transitory computer-readable storage medium comprising program code, wherein when the program code is executed by a computer device, the program code causes the computer device to: (Flynn, [0096] “a computer-accessible medium may include a non-transitory, computer-readable storage medium or memory medium…” Zhou [0124] “an arithmetic coder performs entropy coding….to generate a code stream b1 denoting the image data.” [0125] “…the code stream b1 denoting the image data may be used for network transmission or storage.” )
Claim Rejections - 35 USC § 103
Claims 38-44 and 47-52 are rejected under 35 U.S.C. 103 as being unpatentable over Zhou-Flynn in view of Mokrushin et al (US-10127913-B1) hereinafter “Mokrushin”.
Regarding Claim 38 Zhou-Flynn-Mokrushin
Zhou-Flynn discloses 38. (New) The method according to claim 37,
wherein determining whether the entropy encoding needs to be performed on the first feature element (Zhou, [0196] “…in performing entropy coding on the data to be coded in a channel in the discrete feature maps according to the flag data….when the flag data indicate that the data of a channel in the discrete feature maps are all zero, no entropy coding is performed on the data to be coded of the channel in the discrete feature maps; and when the flag data indicate that data of a channel in the feature maps are not all zero, entropy coding is performed on the data to be coded according to the probabilities of the data to be coded of the channel.” comprises:
obtaining a probability estimation result of the first feature element; (Zhou, [0084] Fig, 1“104: probabilities of to-be-coded data in the discrete feature maps are calculated according to the preprocessed data;”) and
Zhou-Flynn does not expressly disclose
determining, based on the probability estimation result of the first feature element, whether to perform the entropy encoding on the first feature element.
However, in the same field of endeavor Mokrushin discloses more explicitly the following:
determining, based on the probability estimation result of the first feature element, whether to perform the entropy encoding on the first feature element. (Mokrushin, Col. 17, lines 22-26 “If the probability thus obtained is in the range that is predefined with pre-established values and/or the counter of a context occurrence number has a value greater or smaller than the predefined value, then entropy encoding is not carried out,..” i.e., the context model group processor (711) determines, based on the obtained probability, weather entropy encoding should be executed. When the obtained probability satisfies the predefined condition, entropy encoding it is performed.)
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the application to modify the teachings of Zhou in view of Mokrushin to create the system of Zhou-Flynn as outlined above in order to perform the entropy encoding on a first feature element only when it is determined, based on the probability estimation, that the entropy encoding needs to be performed.
One ordinary skill in the art would have been motivated to incorporate Mokrushin’s conditional encoding control into Zhou-Flynn feature data coding system to avoid unnecessary entropy operations when probability thresholds or context occurrence values indicate that entropy coding would have not provide sufficient benefit. Such modification would predictably reduce computational complexity and improve coding efficiency while maintaining coding accuracy. (Mokrushin, Col, 2, lines 41-47)
Regarding Claim 39 Zhou-Flynn-Mokrushin
Zhou-Flynn-Mokrushin 39. (New) The method according to claim 38,
wherein determining, based on the probability estimation result of the first feature element, whether to perform the entropy encoding on the first feature element comprises:
when the probability estimation result of the first feature element meets a preset condition, determining that the entropy encoding needs to be performed on the first feature element; (Mokrushin, Col. 30, lines 8-11 “ the data on the probability is extracted from the selected cell of the selected context model, which data is used for entropy encoding of a current bit of the data stream” Fig. 16, Col. 16, lines 36-41 “If no one of the bypass routes is selected, the obtained bits are transferred to the entropy coder (1616) where current bits are encoded with the use of a probability predicted by the context model in the context modeling system (1604). The encoded data is transferred to the output data stream.” ) and
when the probability estimation result of the first feature element does not meet the preset condition, determining that the entropy encoding does not need to be performed on the first feature element. (Mokrushin, Col. 23, lines 46-56 “If the probability thus obtained is within the range set by predefined values, and/or the context counter has a value which is greater or smaller than a predefined value, then entropy decoding is not carried out, and the value of the bit transfers, through the pass switch for bits of the binarized string (1806) from the input stream to the input of the debinarizer (1816), bypassing the entropy decoder (1815), or, if the binary syntactic element is decoded, the value of this syntactic element is transferred, through the syntactic element pass switch (1805), to the output data stream (1817), bypassing the subsequent step of decoding.” Col. 33, lines 17-24 “wherein at least one syntactic element, depending on a value of at least one context element associated therewith, and/or depending on values of the probability calculated, and/or depending on a value of an individual counter of context occurrence number, is written into the data stream directly, bypassing the step of encoding.”)
Regarding Claim 40 Zhou-Flynn-Mokrushin
Zhou-Flynn-Mokrushin discloses 40. (New) The method according to claim 39, wherein when the probability estimation result of the first feature element is a probability value that a value of the first feature element is k, the preset condition is that the probability value that the value of the first feature element is k is less than or equal to a first threshold, wherein k is an integer, and wherein k is one of a plurality of candidate values of the first feature element. (Zhou, [0084] “probabilities of to-be-coded data in the discrete feature maps are calculated according to the preprocessed data;..” further Mokrushin discloses Col. 17, lines 18-26 “The context model group processor (711) uses values of the counters of a context occurrence number in the corresponding cells of the corresponding context models for selecting the current context model and reading off the probability. If the probability thus obtained is in the range that is predefined with pre-established values and/or the counter of a context occurrence number has a value greater or smaller than the predefined value, then entropy encoding is not carried out…”)
Regarding Claim 41 Zhou-Flynn-Mokrushin
Zhou-Flynn-Mokrushin discloses 41. (New) The method according to claim 40, further comprising:
constructing a threshold candidate list; (Mokrushin teaches using pred-established/predefined values as threshold conditions for determining whether entropy encoding is carried out Col. 17, lines 18-26. It would have been obvious to arrange such predefined threshold values as a threshold candidate list.)
putting the first threshold into the threshold candidate list; (Since Mokrushin uses the predefined threshold/range value as the condition for decoding whether entropy encoding is performed or bypass, it would be obvious to include the selected first threshold in the threshold candidate list.)
writing an index number corresponding to the first threshold into an encoded bitstream, wherein a length of the threshold candidate list is T, and wherein T is an integer greater than or equal to 1. (Flynn teaches explicitly specifying information in the bitstream as header information. and further teaches using a packet index to indicate packet order .Flynn [0071]-[0072] ” a certain number N of bytes (e.g., 1 byte) could be used to indicate the order, which specifies the packet index modulo a number 2{circumflex over ()}N.” Flynn also teaches [0075] “the packet ordering information could be combined with the probability reset information described in the previous section in order to achieve a more compact header. For example, use 1-2 bits for the probability information and the rest (e.g., 7-6 bits) for the ordering information.”)
Regarding Claim 42 Zhou-Flynn-Mokrushin
Zhou-Flynn-Mokrushin discloses 42. (New) The method according to claim 40,
wherein the probability value that the value of the first feature element is k is a maximum probability value in probability values of all candidate values of the first feature element. (Zhou discloses calculating probability of to be-coded in the discrete feature maps ¶ ¶[0084]. Zhou further discloses estimating probabilities of the data to be coded and performing entropy coding according to the probabilities of the to-be-coded data. ( Zhou, ¶¶ [0103]-[0104], ¶¶[0124] since the feature data is discrete, the calculated probability values of the feature element, and it would have been obvious to identify the candidate value k having the maximum probability value among the probability values of all candidate values.)
Regarding Claim 43 Zhou-Flynn-Mokrushin
Zhou-Flynn-Mokrushin discloses 43. (New) The method according to claim 39,
wherein when the probability estimation result of the first feature element comprises a parameter of probability distribution of the first feature element, the preset condition is that the parameter of the probability distribution of the first feature element is greater than or equal to a second threshold. (Zhou discloses that a probability model may be a Gaussian probability or a Laplacian probability model, and that the Gaussian probability model may be obtained by estimating a mean and a variance or standard deviation σ. Zhou, ¶[0091] Zhou further discloses [0124]”…a mean value μ and a standard deviation σ of Gaussian distribution..” Mokrushin teaches making an entropy-coding decision based on calculated probability values and/or predefined values. Mokrushin, Col. 4, lines 10-15 Col. 17, lines 18-26. Thus, it would have been obvious to apply Mokrushin’s predefined threshold condition to Zhou’s probability-distribution parameter, such as the mean or standard deviation, to determine whether entropy encoding needs to be performed.)
Regarding Claim 44 Zhou-Flynn-Mokrushin
Zhou-Flynn-Mokrushin discloses 44. (New) The method according to claim 43, wherein:
when the probability distribution is a Gaussian distribution, the parameter of the probability distribution of the first feature element is a mean value of the Gaussian distribution of the first feature element or a variance of the Gaussian distribution of the first feature element; (Zhou, [0091] “…, in 104, calculation of any type of probability model may be performed; for example, a Gaussian probability model, …the Gaussian probability model as an example, the Gaussian probability model may be obtained by estimating a mean value and variance or standard deviation of Gaussian distribution. [0124] “…taking Gaussian distribution as an example, a probability model includes a mean value μ and a standard deviation σ of Gaussian distribution, thereby obtaining the probabilities … of the data to be coded; a flag data generator generates flag data flag indicating whether the discrete latent image representations ŷ are all zero on channels, and an arithmetic coder performs entropy coding on the discrete latent image representations ŷ according to the probabilities…”) and
when the probability distribution is a Laplace distribution, the parameter of the probability distribution of the first feature element is a location parameter of the Laplace distribution of the first feature element or a scale parameter of the Laplace distribution of the first feature element. (Zhou, [0091] “ calculation of any type of probability model may be performed; for example, a Gaussian probability model, a Laplacian probability model, or the like, may be used”.[0136] “In 602, calculation of any type of probability model may be performed; for example, a Gaussian probability model, a Laplacian probability model, or the like, may be used.”. A person of ordinary skill in the art would have understood that a Laplace/Laplacian distribution is characterized by distribution parameters including a location parameters and a scale parameter.)
Regarding Claim 47 Zhou-Flynn-Mokrushin
Zhou-Flynn-Mokrushin 47. (New) The method according to claim 38,
wherein determining, based on the probability estimation result of the first feature element, whether to perform the entropy encoding on the first feature element comprises:
inputting a probability estimation result of the feature data into a generative network to obtain decision information of the first feature element; (Zhou, Fig. 11 “convolutional neural network coder 1101” [0186] a convolutional neural network coder 1101 configured to perform feature extraction on to-be-processed image data by using a convolutional neural network, so as to generate feature maps of the image data;” [0198] “…,the probability estimator 1104 may further calculate the probabilities of the data to be coded in the discrete feature maps according to the flag data generated by the data generator) and
determining, based on the decision information of the first feature element, whether to perform the entropy encoding on the first feature element. (Zhou, Fig. 11 [0190] “…an entropy coder 1105 configured to perform entropy coding on the to-be-coded data according to the probabilities of the to-be-coded data” [0196] “…when the flag data indicate that data of a channel in the discrete feature maps are not all zero, the entropy coder performs entropy coding on the to-be-coded data in the channel according to the probabilities of the to-be-coded data.”)
Regarding Claim 48 Zhou-Flynn-Mokrushin
Zhou-Flynn-Mokrushin discloses 48. (New) The method according to claim 47, wherein the decision information of the first feature element is a decision map, and wherein determining, based on the decision information of the first feature element, whether to perform the entropy encoding on the first feature element comprises:
when a value corresponding to a location at which the first feature element is located in the decision map is a preset value, determining that the entropy encoding needs to be performed on the first feature element; (Zhou, Fig. 11 [0190] “…an entropy coder 1105 configured to perform entropy coding on the to-be-coded data according to the probabilities of the to-be-coded data” [0196] “…when the flag data indicate that data of a channel in the discrete feature maps are not all zero, the entropy coder performs entropy coding on the to-be-coded data in the channel according to the probabilities of the to-be-coded data.”) and
when the value corresponding to the location at which the first feature element is located in the decision map is not the preset value, determining that the entropy encoding does not need to be performed on the first feature element. (Zhou, [0196] “when the flag data indicate that data of a channel in the discrete feature maps are all zero, the entropy coder 1105 does not perform entropy coding on to-be-coded data in the discrete feature maps of the channel.”)
Regarding Claim 49 Zhou-Flynn-Mokrushin
Zhou-Flynn-Mokrushin discloses 49. (New) The method according to claim 47, wherein determining, based on the decision information of the first feature element, whether to perform the entropy encoding on the first feature element comprises:
when the decision information of the first feature element is a preset value, determining that the entropy encoding needs to be performed on the first feature element; (Zhou, Fig. 11 [0190] “…an entropy coder 1105 configured to perform entropy coding on the to-be-coded data according to the probabilities of the to-be-coded data” [0196] “…when the flag data indicate that data of a channel in the discrete feature maps are not all zero, the entropy coder performs entropy coding on the to-be-coded data in the channel according to the probabilities of the to-be-coded data.”) and
when the decision information of the first feature element is not the preset value, determining that the entropy encoding does not need to be performed on the first feature element. (Zhou, [0196] “when the flag data indicate that data of a channel in the discrete feature maps are all zero, the entropy coder 1105 does not perform entropy coding on to-be-coded data in the discrete feature maps of the channel.”)
Regarding Claim 50 Zhou-Flynn-Mokrushin
Zhou-Flynn- Mokrushin discloses 50. (New) The method according to claim 37, wherein determining whether the entropy encoding needs to be performed on the first feature element comprises:
obtaining side information of the feature data; (Zhou, [0158] “901: feature extraction is performed on to-be-processed image data by using a convolutional neural network, to generate feature maps of the image data; ”)
inputting the side information of the feature data into a joint network to obtain decision information of the first feature element; (Zhou, [0243] “…the processor 1510 may be configured to perform following control: performing feature extraction on to-be-processed image data by using a convolutional neural network, so as to generate feature maps of the image data; quantizing the feature maps to generate discrete feature maps; preprocessing the discrete feature maps to generate preprocessed data…” ) and
determining, based on the decision information of the first feature element, whether to perform the entropy encoding on the first feature element. (Zhou, [0243] “.. feature maps to generate preprocessed data, an amount of data of the preprocessed data being less than an amount of data of the discrete feature maps; calculating probabilities of to-be-coded data in the discrete feature maps according to the preprocessed data; and performing entropy coding on the to-be-coded data according to the probabilities of the to-be-coded data.” Mokrushin further teaches determining whether entropy encoding is performed or bypassed based on probability related information or predefined conditions. Mokrushin, Col. 17, lines 18-26)
Regarding Claim 51 Zhou-Flynn-Mokrushin
Zhou-Flynn discloses 51. (New) The method according to claim 50,
wherein the decision information of the first feature element is a decision map, and wherein determining, based on the decision information of the first feature element, whether to perform the entropy encoding on the first feature element comprises:
when a value corresponding to a location at which the first feature element is located in the decision map is a preset value, determining that the entropy encoding needs to be performed on the first feature element; (Zhou discloses performing feature extraction on to-be-encoded image data using a convolution neural network to generate feature maps of the image data, quantizing the feature maps to generate discrete feature maps, preprocessing the discreet feature maps, calculating probabilities of to-be-coded data according to the preprocessed data, and performing entropy coding according to the probabilities. Zhou, [0243] ) and
when the value corresponding to the location at which the first feature element is located in the decision map is not the preset value, determining that the entropy encoding does not need to be performed on the first feature element. (Zhou, [0196] “when the flag data indicate that the data of a channel in the discrete feature maps are all zero, no entropy coding is performed on the data to be coded of the channel in the discrete feature maps; Thus, the flag/decision value corresponding to all zero condition teaches the non-preset value branch in which entropy is not performed. Mokrushin further teaches Col. 17, lines 22-26 “if the probability thus obtained is in the range that is predefined with pre-established values…, then entropy encoding is not carried out,…”)
Regarding Claim 52 Zhou-Flynn-Mokrushin
Zhou-Flynn-Mokrushin discloses 52. (New) The method according to claim 50, wherein determining, based on the decision information of the first feature element, whether to perform the entropy encoding on the first feature element comprises:
when the decision information of the first feature element is a preset value, determining that the entropy encoding needs to be performed on the first feature element; (Zhou, [0167] “when the flag data indicate that data of a channel in the feature maps are not all zero, entropy coding is performed on the data to be coded according to the probabilities of the data to be coded of the channel.) and
when the decision information of the first feature element is not the first preset value, determining that the entropy encoding does not need to be performed on the first feature element. (Zhou, [0167] “when the flag data indicate that the data of a channel in the discrete feature maps are all zero, no entropy coding is performed on the data to be coded of the channel in the discrete feature maps;” Mokrushin further teaches using predefined values/conditions to determine whether entropy encoding is performed or bypassed. Mokrushin Col. 17, lines 18-26.)
Claim Rejections - 35 USC § 103
Claim 45 is rejected under U.S.C. 103 as being unpatentable over Zhou-Flynn-Mokrushin in view of Przyborowski et al (US-20250181943-A1) hereinafter “Przyborowski”.
Regarding Claim 45 Zhou-Flynn-Mokrushin-Przyborowski
Zhou-Flynn-Mokrushin discloses 45. (New) The method according to claim 39,
Zhou-Flynn-Mokrushin does not expressly disclose
wherein the probability estimation result of the first feature element is obtained through Gaussian mixture distribution, and wherein the preset condition is:
a sum of any variance of the Gaussian mixture distribution of the first feature element and a sum of absolute values of differences between all mean values of the Gaussian mixture distribution of the first feature element and a value k of the first feature element is greater than or equal to a fifth threshold;
a difference between any mean value of the Gaussian mixture distribution of the first feature element and a value k of the first feature element is greater than or equal to a sixth threshold; or any variance of the Gaussian mixture distribution of the first feature element is greater than or equal to a seventh threshold, wherein k is an integer, and wherein k is one of a plurality of candidate values of the first feature element.
However, in the same field of endeavor Przyborowski discloses more explicitly the following:
wherein the probability estimation result of the first feature element is obtained through Gaussian mixture distribution, (Przyborowski, [0016] “…a method and/or devices of an efficient Gaussian Mixture Model (GMM) distribution based approximation of a data set in a computing environment..”) and wherein the preset condition is:
a sum of any variance of the Gaussian mixture distribution of the first feature element and a sum of absolute values of differences between all mean values of the Gaussian mixture distribution of the first feature element and a value k of the first feature element is greater than or equal to a fifth threshold; (Przyborowski [0027] …, in one or more embodiments, the EM algorithm may include two operations, viz. expectation and maximization. Assuming that the GMM/GMM distribution 180 includes q components or q Gaussian distributions therein, in one or more embodiments, as part of the expectation operation, for each observation x.sub.j∈{x.sub.1, x.sub.2 . . . x.sub.l}, the probability that x.sub.j originates from the i.sup.th (i={1, 2 . . . q}) Gaussian distribution may be computed. In one or more embodiments, as part of the maximization operation, for each i={1, 2 . . . q}, parameters μ.sub.i, Σ.sub.i and w.sub.i (μ.sub.i may refer to a mean of the i.sup.th Gaussian distribution, Σ.sub.i may refer to a variance of the i.sup.th Gaussian distribution for a univariate form thereof or a covariance (e.g., in matrix form) of the i.sup.th Gaussian distribution for a multivariate form thereof, and w.sub.i may refer to a weight of the i.sup.th Gaussian distribution indicative of a probability that input data 170 x.sub.j belongs to the i.sup.th Gaussian distribution) that maximize an evidence lower bound for all x.sub.j given the aforementioned derived probabilities may be found. …, initial values for the mean of the i.sup.th Gaussian distribution may be obtained through a random guess and/or a heuristic approach such as one involving a k-means (or, q-means) clustering algorithm.”)
a difference between any mean value of the Gaussian mixture distribution of the first feature element and a value k of the first feature element is greater than or equal to a sixth threshold; (Przyborowski, [0031] “…operation 204 may involve modifying input data 170 (e.g., the modified input data 170 is shown as modified data 196 in FIG. 1) by replacing, for each constituent Gaussian distribution of GMM distribution 180, numeric value(s) and/or vector(s) of the numeric values of input data 170 that differ in magnitude from a center of the each constituent Gaussian distribution by less than a threshold value (e.g., threshold 194 shown stored in memory 114.sub.1) with a mean value of input data 170, with a weight of the mean value being indicative of a cardinality (or frequency)… [0034] “criteria 302 may involve the numeric value(s) and/or the vector(s) of the numeric values of input data 170 being different in magnitude from the center of the each constituent Gaussian distribution by less than a numeric distance (e.g., threshold 194) and from the centers of other constituent Gaussian distributions (or, at least one other constituent Gaussian distribution) of GMM distribution 180 by more than another numeric distance (e.g., threshold 310)”) or any variance of the Gaussian mixture distribution of the first feature element is greater than or equal to a seventh threshold, wherein k is an integer, and wherein k is one of a plurality of candidate values of the first feature element.
Therefore, it would have been obvious to a person having ordinary skill in the art before
the effective filing date of the application to modify the teachings of Zhou-Mokrushin with Przyborowski to create the system of Zhou-Mokrushin as outlined above in order to incorporate a first sum of variance and a second sum of absolute values-mean differences of the Gaussian mixture distribution-each compared against respective threshold as suggested by Przyborowski.
The motivation is “for improved efficiency and/or accuracy in the generation of GMM distribution from input data ” (Przyborowski,[0023])
Claim Rejections - 35 USC § 103
Claim 46 is rejected under U.S.C. 103 as being unpatentable over Zhou-Mokrushin in view of Smolak et al (US-9903803-B2) hereinafter “Smolak”.
Regarding Claim 46 Zhou-Flynn-Mokrushin-Smolak
Zhou-Flynn-Mokrushin discloses 46. (New) The method according claim 39,
Zhou-Flynn-Mokrushin does not explicitly disclose
wherein the probability estimation result of the first feature element is obtained through asymmetric Gaussian distribution, and wherein the preset condition is:
an absolute value of a difference between a mean value of the asymmetric Gaussian distribution of the first feature element and a value k of the first feature element is greater than or equal to an eighth threshold;
a first variance of the asymmetric Gaussian distribution of the first feature element is greater than or equal to a ninth threshold; or a second variance of the asymmetric Gaussian distribution of the first feature element is greater than or equal to a tenth threshold, wherein k is an integer, and wherein k is one of a plurality of candidate values of the first feature element.
However, in the same field of endeavor Smolak discloses more explicitly the following:
wherein the probability estimation result of the first feature element is obtained through asymmetric Gaussian distribution, and wherein the preset condition is:
an absolute value of a difference between a mean value of the asymmetric Gaussian distribution of the first feature element and a value k of the first feature element is greater than or equal to an eighth threshold; (Smolak, Col. 9, lines 45-49 Step 354 in which a batch-specific signal peak threshold is determined as a function of the batch-specific noise characteristic may involve setting the signal peak threshold based on the mean μ of the asymmetric Gaussian distribution fit plus an increment.” i.e., threshold are derived based on mean value)
a first variance of the asymmetric Gaussian distribution of the first feature element is greater than or equal to a ninth threshold; (Smolak, Col. 9, lines 50-55 “…peak threshold may be set to be the greater of the mean μ of the asymmetric Gaussian distribution fit plus three times the standard deviation σ of the asymmetric Gaussian distribution fit or the mean μ of the asymmetric Gaussian distribution …, whichever is greater (e.g., signal peak threshold is no smaller,,,) Smolak, Col. 10, lines 31-34” The asymmetric Gaussian distribution fit is also shown referenced by numeral 642, and integer multiples of the standard deviation σ away from the mean μ are indicated by the vertical dashed lines…” i.e., threshold are derived based on mean variance/standard deviation) or a second variance of the asymmetric Gaussian distribution of the first feature element is greater than or equal to a tenth threshold, wherein k is an integer, and wherein k is one of a plurality of candidate values of the first feature element.
Therefore, it would have been obvious to a person having ordinary skill in the art before
the effective filing date of the application to modify the teachings of Zhou-Mokrushin in view of Smolak to create the system of Zhou-Mokrushin as outlined above in order to incorporate the use of an asymmetric distribution element for future element, wherein the thresholds are determined based on the mean and variance of the asymmetric Gaussian distribution
as taught by Smolak.
One of ordinary skill in the art would have been motivated to incorporate the asymmetric Gaussian distribution into the system Zhou-Mokrushin such that the “efficiency of the probability estimation may be improved.”(Zhou,[0090])
Pertinent Prior Art
The prior art made of record and not relied upon is considered pertinent to applicant's
disclosure.
Chelian et al. US-10133983-B1
McNair et al. US-10446273-B1
Wang et al. US-9972314-B2
Oboukhov et al. US-11811425-B2
PIAO et al. US-20240283935-A1
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ASTEWAYE GETTU ZEWEDE whose telephone number is (703)756-1441. The examiner can normally be reached Mo-Fr 8:30 am to 5:30 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, William Vaughn can be reached at (571)272-3922. 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.
/ASTEWAYE GETTU ZEWEDE/Examiner, Art Unit 2481
/JERRY T JEAN BAPTISTE/Primary Examiner, Art Unit 2481
Prosecution Insights