DETAILED ACTION
This action is responsive to the Amendments and Remarks received 07/23/2025 in which claims 2, 4–6, 10, 12, 13, and 17–19 are cancelled, claims 1, 9, and 16 are amended, and no claims are added as new claims.
Response to Arguments
Examiner incorporates herein previous Responses to Arguments.
On page 6 of the Remarks, Applicant takes issue with Examiner’s findings from the preceding Office Action, arguing that Applicant’s Specification does not support what Examiner found was being disclosed. Examiner disagrees with or is confused by Applicant’s arguments. Applicant avers on page 6 of Applicant’s Remarks that, “‘the number of frequencies’ refers to the number of non-zero transform coefficients within the current block” and Applicant bases that argument on the fact that paragraph [0321] states, “after transforming a block having a size as shown in FIG. 12, four coefficients located in the upper left may be defined as the lowest frequency as shown in FIG. 13….” Applicant then reproduces FIG. 12 at the bottom of page 6 of the Remarks and FIG. 13 at the top of page 7 of the Remarks. The 2x2 upper-left region in both Figs. 12 and 13 have a zero-valued coefficient! So, Examiner disagrees it necessarily means non-zero valued coefficients. Now, if the argument is that the number of frequencies includes zero valued and non-zero valued coefficients, then size and number of frequencies is the same thing. Does the size of a vector or array mean the same thing as the number of entries in the vector or array? Yes. This argument does not seem to be beneficial to advancing prosecution, so it appears unnecessary to debate it further. Furthermore, Chen’s paragraph [0058] explains it was prior art to count the number of non-zero coefficients to decide whether to use a non-separable secondary transform.
On pages 7–10 of the Remarks, as best Examiner can tell, Applicant seems to be trying to respond to a request for information from the Office. Examiner thanks Applicant for the attempt to be helpful. However, the only bit of helpful information Examiner could glean from Applicant’s explanation is that certain transforms do not perform as well as others under certain conditions leading one to desire to use a different transform. Examiner believes Applicant is trying to explain that some transforms reduce the number of non-zero coefficients better than others. Well, that is why the prior art supports utilizing several different transforms and adaptively selecting the best transform for the block. The art utilizes a coding tool that goes by several equivalent names, such as multi-transform selection (MTS) or adaptive multi-transform (AMT).
On page 11 of the Remarks, Applicant contends the prior art is deficient because it does not teach or suggest selecting from predefined matrices a first and second transform matrix having different coefficient values and insists the transforms be either DCT-2, DCT-4 or DST-4. Examiner finds Applicant does not argue that which is claimed and resists Applicant’s invitation to import limitations from the Specification into the claims. As explained in the rejection, infra, Chen and Jang, in combination, teach or suggest using a secondary transform index for signaling which type of secondary transform among a plurality of available secondary transforms comprising DCT, DST, KLT, etc. Indeed, Examiner notes that the default secondary transform is DCT-2, but other secondary transforms are possible as evidenced by Alshina, cited under the Conclusion Section of this Office Action. If necessary to sustain the rejection, the teachings of Alshina should be considered along with the other prior art of record as an alternative ground of rejection. Further still, Examiner notes that DCT-8 and DST-7 are known for their advantageous use with compacting directional (e.g. horizontal or vertical) non-zero values. In most scenarios, such directionality is less prominent in the reduced, low-frequency coefficient region resultant from the primary transform and therefore DCT-8 and DST-7 as secondary transforms are not expected to be of enough benefit to justify increased complexity of implementation.
On page 12 of the Remarks, Applicant makes a confusing argument about size and Ikai’s failure to disclose such by arguing the claim recites first and second transform matrices. Examiner finds the argument regarding size does not match the averred language. Therefore, despite Applicant’s insistence to the contrary, the amendments to claim 1 do not clarify any difference. And, as explained in the preceding Office Action, Chen’s teachings are principally relied upon for teaching the size constraint feature recited in the claims.
Other claims are not argued separately. Remarks, 12.
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 of this title, 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, 3, 7–9, 11, and 14–16 are rejected under 35 U.S.C. 103 as being unpatentable over Ikai (US 2020/0213626 A1), Chen (US 2021/0120269 A1), and Jang (US 2019/0356915 A1).
Regarding claim 1, the combination of Ikai, Chen, and Jang teaches or suggests a method of decoding a sequence of pictures, the method comprising: decoding, from a sequence parameter set referenced by the sequence of pictures, information on a maximum block size for which a secondary inverse transform is allowed (Ikai, Abstract: teaches a secondary transform can be omitted in accordance with a size of the current block; Ikai, ¶ 0087: teaches maximum block sizes are defined in the SPS; While Ikai teaches conditioning the use of secondary transforms on block size, Jang more explicitly teaches why it would be desirable to set a maximum block size for secondary transforms; Jang, ¶ 0206: teaches that applying secondary transforms to larger block sizes increases complexity and can degrade compression performance, thus suggesting to the skilled artisan that one could benefit from the ability to set a maximum block size for which a secondary transform is allowed; Examiner notes the skilled artisan in this art knows all too well that constraints like block size constraints can be set in the SPS); performing dequantization on a current block to obtain first transform coefficients of the current block (Ikai, ¶ 0139: teaches inverse quantization performed at a decoder; see also Fig. 2: illustrating the functional blocks of a decoder; Examiner notes Ikai teaches the corresponding (opposite) encoder operations; see e.g. Ikai, ¶ 0284); performing a secondary inverse transform on the first transform coefficients of the current block to obtain second transform coefficients of the current block (Ikai, ¶ 0139: teaches inverse transform; Ikai, ¶ 0011: teaches first and second transforms; see also Ikai, ¶ 0237: describing the first and second transforms as “core” and “secondary”; see e.g. Ikai, ¶ 0284: teaching the correct order of operations for the encoder and decoder); performing a primary inverse transform on the second transform coefficients of the current block to obtain a residual block of the current block (Ikai, ¶¶ 0114–0116: explains the transform unit includes a quantized predictive residual, which means the encoder subtracts a prediction from the original image to obtain a residual and then the residual is transformed and quantized; The decoder works in the opposite direction, first inverse quantizing and inverse transforming to obtain the residual to be added back to the prediction; see also Fig. 2: illustrating the functional blocks of a decoder; see e.g. Ikai, ¶ 0284: teaching the correct order of operations for the encoder and decoder); and adding the residual block of the current block and a prediction block of the current block to obtain a reconstructed block of the current block (see description, supra, regarding residuals and prediction; see also e.g. Ikai, ¶ 0135), wherein the secondary inverse transform is performed only when the current block is intra-predicted (Ikai, ¶ 0213: teaches “The Secondary Transform is a transform selected in the intra-prediction mode.”; see also, Ikai, ¶ 0477), wherein a first number of the first transform coefficients on which the secondary inverse transform is performed is determined based on a size of the current block (Chen, ¶ 0020: teaches the size of the block informs the size of the secondary transform; large blocks are more likely to have the 8x8 secondary transform applied, although 4x4 might also be best even for large block sizes), wherein the secondary inverse transform is not performed when the size of the current block is larger than the maximum block size (Jang, ¶ 0206: teaches that applying secondary transforms to larger block sizes increases complexity and can degrade compression performance, thus suggesting to the skilled artisan that one could benefit from the ability to set a maximum block size for which a secondary transform is allowed; Ikai, ¶ 0444: teaches the secondary transform can be constrained to being only performed when the size of the transform block is not larger than a prescribed size; while not relied upon, it is noted that Zhao cited under the Conclusion Section of this Office Action also teaches this feature) , and wherein performing the secondary inverse transform comprises: selecting, from predefined transform matrices including at least a first transform matrix and a second transform matrix which have different coefficient values, a transform matrix of the secondary inverse transform based on an intra prediction mode and the size of the current block (Ikai, ¶¶ 0315–0322 and Fig. 54A: teaches the secondary transform is determined by the intra-prediction mode wherein a transform matrix is indicated by a host of parameters matching those of the core transform, but for the second transform; Examiner notes selecting a transform is selecting a transform matrix (kernel in previous version of the claim); Ikai, Abstract: teaches a secondary transform can be omitted in accordance with a size of the current block; see also Ikai, ¶¶ 0372 and 0374; see also Chen, ¶¶ 0016, 0019, and 0020: teaches the many transform matrices having different coefficient values that are possible for a secondary transform wherein secondary transform candidates are predefined according to intra-prediction mode and size of the current block; Chen, ¶ 0022: teaches, “A secondary transform is then selected and applied….”; According to Applicant’s Remarks, dated 02/13/2025, it appears there is some debate whether the secondary transform can be selected from a group consisting of the known variations of DCT and DST used in the art; Examiner does not find it reasonable to contend the secondary transform cannot be one of the many types, but as a dictionary-type reference for understanding what the prior art is teaching a skilled artisan, Examiner draws attention to Alshina, cited under the Conclusion Section of this Office Action, which explains the secondary transform can be based on versions of DCT and/or DST); performing the secondary inverse transform on the first transform coefficients using the first transform matrix of the secondary inverse transform when the first transform matrix is selected (Ikai, ¶ 0139: teaches inverse transform; Ikai, ¶ 0011: teaches first and second transforms; see also Ikai, ¶ 0237: describing the first and second transforms as “core” and “secondary”; see e.g. Ikai, ¶ 0284: teaching the correct order of operations for the encoder and decoder; Examiner notes the primary and secondary terminology is from the standpoint of the encoder and when discussing the corresponding decoder operations, the inverse transforms are applied in the order of the secondary transform being performed first and the primary transform being performed second; Chen, ¶¶ 0016, 0019, and 0020: teaches the many transform matrices having different coefficient values that are possible for a secondary transform wherein secondary transform candidates are predefined according to intra-prediction mode and size of the current block; JChen, ¶ 0058: teaches an NSST index for signaling the secondary transform; Jang, ¶ 0236: teaches secondary inverse transforms can be selected from among DCT, DST, KLT, etc.; In combination then, Chen and Jang teach or suggest using a secondary transform index for signaling which type of secondary transform among a plurality of secondary transforms to select); and performing the secondary inverse transform on the first transform coefficients using the second transform matrix of the secondary inverse transform when the second transform matrix is selected (Chen, ¶¶ 0016, 0019, and 0020: teaches the many transform matrices having different coefficient values that are possible for a secondary transform wherein secondary transform candidates are predefined according to intra-prediction mode and size of the current block; Chen, ¶ 0058: teaches an NSST index for signaling the secondary transform; Jang, ¶ 0236: teaches secondary inverse transforms can be selected from among DCT, DST, KLT, etc.; In combination then, Chen and Jang teach or suggest using a secondary transform index for signaling which type of secondary transform among a plurality of secondary transforms to select).
One of ordinary skill in the art, before the effective filing date of the claimed invention, would have been motivated to combine the elements taught by Ikai, with those of Chen, because both references are drawn to the same field of endeavor (Secondary Transforms according to e.g. size) such that one wishing to practice the art of secondary transforms for video coding would be led to their relevant teachings, and because, as Chen explains, a 4x8 or 8x4 block partition could not make use of a 8x8 transform kernel because it is too large along one dimension. Thus, the combination represents a mere combination of prior art elements, according to known methods, to yield a predictable result. This rationale applies to all combinations of Ikai and Chen used in this Office Action unless otherwise noted.
One of ordinary skill in the art, before the effective filing date of the claimed invention, would have been motivated to combine the elements taught by Ikai and Chen, with those of Jang, because all three references are drawn to the same field of endeavor (Secondary Transforms according to e.g. size) such that one wishing to practice the art of secondary transforms for video coding would be led to their relevant teachings, and because, as Jang explains, applying secondary transforms to larger block sizes adds undesirable complexity and could degrade compression performance (Jang, ¶ 0206). Thus, the combination represents a mere combination of prior art elements, according to known methods, to yield a predictable result. This rationale applies to all combinations of Ikai, Chen, and Jang used in this Office Action unless otherwise noted.
Regarding claim 3, the combination of Ikai, Chen, and Jang teaches or suggests the method of claim 1, wherein the secondary inverse transform is performed using a low frequency inverse transform (Ikai, ¶ 0506: “In addition, the Secondary Transform unit 1522 may be configured to apply the Secondary Transform to: a first region including a sub-blocks (sic) located on the lower frequency side of the TU;”).
Regarding claim 7, the combination of Ikai and Chen teaches or suggests the method of claim 1, wherein the secondary inverse transform is performed after rearranging the first transform coefficients of the current block from a 2D block format to a 1D list format (Ikai, ¶¶ 0297–0298: teach the 4x4 sub-block is transformed using a Secondary Transform wherein the secondary transform is performed on a one-dimensional vector, i.e. the 4x4 2D matrix is converted to a 16x1 vector for the secondary transform).
Regarding claim 8, the combination of Ikai and Chen teaches or suggests the method of claim 1, wherein the secondary inverse transform is performed in an application range determined on the basis of a smaller value of a width or a height of the current block (Chen, ¶ 0020: teaches that if either the width or the height is the smaller size of 4, then the 4x4 NSST is used rather than the 8x8).
Claim 9 lists essentially the same elements as claim 1, but is drawn to the corresponding encoding method. Therefore, the rationale for the rejection of claim 1 applies to the instant claim.
Claim 11 lists essentially the same elements as claim 3, but is drawn to the corresponding encoding method. Therefore, the rationale for the rejection of claim 3 applies to the instant claim.
Claim 14 lists essentially the same elements as claim 7, but is drawn to the corresponding encoding method. Therefore, the rationale for the rejection of claim 7 applies to the instant claim.
Claim 15 lists essentially the same elements as claim 8, but is drawn to the corresponding encoding method. Therefore, the rationale for the rejection of claim 8 applies to the instant claim.
Claim 16 lists essentially the same elements as claim 1, but is a product-by-process claim drawn to a CRM creating using the encoding method, the encoding method related to the decoding method of claim 1. Therefore, the rationale for the rejection of claim 1 applies to the instant claim.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Saxena et al., “On Secondary Transforms for Intra Prediction Residual,” 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), March 2012.
Tao (US 2014/0327759 A1), which explains DCT-II is the most common transform (¶ 0390).
Lee (US 2013/0101048 A1), which explains DCT-II “is a popular compression structure and accepted as the best suboptimal transformation….” (¶ 0004)
Stipanovich (US 2009/0196185 A1), paragraph [0072], explains, “DCT-II is probably the most commonly used form, and is often simply referred to as ‘the DCT.’”
Chen (US 6,658,161 B1) teaches, “DCT-II is the most popular.”
Lim (US 2019/0356913 A1) teaches defining a min and max block size at the SPS (¶ 0044).
Alshina (US 2018/0309990 A1) teaches the secondary transform can be just like the primary transform and can be based on DCT and/or DST, but on fewer coefficients than the primary transform (¶¶ 0077 and 0083).
Zhao (US 2017/0094314 A1) teaches that when the block is greater than or less than a pre-defined threshold, signaling of the NSST index may be skipped and no secondary transform is applied (¶ 0168).
THIS ACTION IS MADE FINAL. 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 mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Michael J Hess whose telephone number is (571)270-7933. The examiner can normally be reached Mon - Fri 9:00am-5:30pm.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, William Vaughn can be reached on (571)272-3922. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8933.
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.
/MICHAEL J HESS/Examiner, Art Unit 2481