DETAILED ACTION
Claims 1-20 are pending for examination.
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 .
Information Disclosure Statement
The information disclosure statement (IDS) was submitted on 08/28/2025. The
submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information
disclosure statement is being considered by the examiner.
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, 3, 6 and 7 are rejected under 35 U.S.C. 103 as being unpatentable over Lin et al. "CE.3-1.7: Multiple-model LM with small block size restriction" Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, 14th Meeting: Geneva, CH, 19-27 March 2019 (Park), in view of Park, US 20190340502 A1 (Park).
Regarding Claim 1, Lin discloses a method comprising: obtaining video data, including data representing a video data region (Lin, pg. 1, section 1– MMLM is similar to CCLM but applies more than 1 model to predict a chroma block);
and computing models, used for cross-component based prediction of chroma samples from the video data region, the computing comprises (Lin, pg. 1, section 1– MMLM is similar to CCLM but applies more than 1 model to predict a chroma block): accumulating data of a first model and of a second model through reference samples selected from the video data (Lin, pg. 1, section 1– For the multi-model derivation, the average of the luma reference samples is used to classify all the luma reference samples into two groups. Then, a linear model is derived for each group…).
However, Lin does not explicitly disclose concurrently using a first loop.
Park teaches concurrently using a first loop (Park ¶ [0094]– Thus, two independent convolution operations are executed in parallel by the neural engines 314A and 314B in a single loop).
Therefore, it 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 to modify Lin to incorporate accumulating data currently for two models in a loop as taught by Park.
One would be motivated to combine Park’s incorporation of accumulating data currently for two models to achieve data accumulation in a single loop (Park ¶ [0094]– Thus, two independent convolution operations are executed in parallel by the neural engines 314A and 314B in a single loop).
Regarding Claim 3, Lin and Park teach the method of Claim 1, as outlined above. In addition, Lin discloses further comprising: applying the first model and the second model through samples of the video region, wherein the first model and the second model are applied to predict the chroma samples (Lin, pg. 1, section 1– MMLM… applies more than 1 model to predict a chroma block) according to their classification into a first class or a second class (Lin, pg. 1, section 1– For the multi-model derivation, the average of the luma reference samples is used to classify all the…samples into two groups).
However, Lin does not explicitly disclose using a second loop.
Park teaches using a second loop (Park [0065]– The second...loop performs convolution operation for each slice in the input data).
Therefore, it 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 to modify Lin to incorporate a second loop as taught by Park.
One would be motivated to combine Park’s incorporation of a second loop for further filtering (Park [0065]– The second...loop performs convolution operation for each slice in the input data).
Regarding Claim 6, Lin, in combination, further discloses the method of Claim 1, wherein the accumulated data of the first model and of the second model are generated (Lin, pg. 1, section 1– For the multi-model derivation, the average of the luma reference samples is used to classify all the...reference samples into two groups. Then, a linear model is derived for each group…) based on the reference samples according to their respective classification into a first class and a second class (Lin, pg. 1, section 1– For the multi-model derivation, the average of the luma reference samples is used to classify all the…samples into two groups).
Regarding Claim 7, Lin and Park teach the method of Claim 6, as outlined above. In addition, Lin discloses further comprising: extracting, based on the reference samples, one or more features (i.e., an average), used for the classification of the reference samples into the first class and the second class (Lin, pg. 1, section 1– For the multi-model derivation, the average of the luma reference samples is used to classify all the luma reference samples into two groups).
However, Lin does not explicitly disclose using the first loop through the reference samples.
Park teaches using the first loop through the reference samples (Park [0094]– Thus, two independent convolution operations are executed in parallel by the neural engines 314A and 314B in a single loop).
Therefore, it 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 to modify Lin to incorporate using a first loop through the reference samples as taught by Park.
One would be motivated to combine Park’s incorporation of accumulating data currently for two models to achieve data accumulation in a single loop. Please see motivation of Claim 1.
Claim 2 is rejected under 35 U.S.C. 103 as being unpatentable over Lin, in view of Park, and further in view of Henderson, US 20210141799 A1 (Henderson).
Regarding Claim 2, Lin and Park teach the method of Claim 1, as outlined above. In addition, Lin discloses of the first model and the second model (Lin, pg. 1, section 1– the average of the luma reference samples is used to classify all the luma reference samples into two groups. Then, a linear model is derived for each group…).
However, Lin does not explicitly disclose rescaling at least one parameter, wherein the rescaling includes rescaling the parameter into a lower bit size representation.
Henderson teaches rescaling at least one parameter, wherein the rescaling includes rescaling the parameter into a lower bit size representation (Henderson [0237]– The first and second model are compressed using subword-level parameterisation and quantisation. Quantization of the stored embeddings, as well as optionally that of other neural network parameters, reduces model storage requirements).
Therefore, it 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 to modify Lin to incorporate rescaling a parameter into a lower bit size representation as taught by Henderson.
Although Henderson does not expressly state a lower bit-size representation, a person of ordinary skill in the art would understand quantization to mean reducing the number of bits used to represent the parameters, thereby providing a lower bit-size representation. One would be motivated to combine Henderson’s incorporation of rescaling a parameter into a lower bit size representation to increase memory efficiency (Henderson ¶ [0030]– Having a smaller model in terms of the number of parameters and storage required means that the model is more memory efficient and can be stored on small devices, e.g., mobile phones or tablets, with limited memory storage for example).
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Lin, in view of Park, and further in view of Pednault et al, US 20030176931 A1 (Pednault).
Regarding Claim 4, Lin and Park teach the method of Claim 1, as outlined above.
However, Lin does not explicitly disclose further comprising: combining the accumulated data of the first model and of the second model into a combined data of a single model.
Pednault teaches further comprising: combining the accumulated data of the first model and of the second model into a combined data of a single model (Pednault ¶ [0429]– update the sufficient statistics stored in the model object by combining the existing sufficient statistics with those of the input model object, and to then construct a new model based on the updated sufficient statistics).
Therefore, it 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 to modify Lin to incorporate combining data of two models into one as taught by Pednault.
One would be motivated to combine Pednault’s incorporation of combining data of two models into one to attain high levels of predictive accuracy (Pednault ¶ [0017]– models afford the flexibility needed to attain high levels of predictive accuracy…).
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over Lin, in view of Park, and further in view of Pednault and Henderson.
Regarding Claim 5, Lin, Park and Pednault teach the method of Claim 4, as outlined above.
However, Lin in view of Park and Pednault does not explicitly disclose further comprising: rescaling at least one parameter, wherein the rescaling includes rescaling the parameter into a lower bit size representation.
Henderson teaches further comprising: rescaling at least one parameter, wherein the rescaling includes rescaling the parameter into a lower bit size representation (Henderson ¶ [0237]– The first and second model are compressed using subword-level parameterisation and quantisation. Quantization of the stored embeddings, as well as optionally that of other neural network parameters, reduces model storage requirements).
Therefore, it 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 to modify Lin to incorporate rescaling a parameter into a lower bit size representation as taught by Henderson.
One would be motivated to combine Henderson’s incorporation of rescaling a parameter into a lower bit size representation to increase memory efficiency. Please see the motivation of Claim 2.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Lin, in view of Park, and further in view of Blanch et al, GB 2591806 A (Blanch).
Regarding Claim 8, Lin and Park teach the method of Claim 7, as outlined above.
However, Lin does not explicitly disclose wherein the extracted one or more features are extracted based on a subset of the reference samples.
Blanch teaches wherein the extracted one or more features are extracted based on a subset of the reference samples (Blanch col. 4, lines 7-12– The first two branches work concurrently to extract features from the available reconstructed samples, including the already reconstructed luma block as well as the neighbouring luma and chroma reference samples. The first branch (referred to as cross-component boundary branch) aims at extracting cross-component information from neighbouring reconstructed samples, using an extended reference array on the left of and above the current block).
Therefore, it 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 to modify Lin to incorporate extracting features based on a subset of reference samples as taught by Blanch.
Although Blanch does not explicitly state that features are extracted based on a subset of the reference samples, Blanch teaches extracting features from an extended reference array comprising neighbouring reconstructed luma and chroma reference samples that surround the current block. A person having ordinary skill in the art would understand that this extended reference array represents only a selected portion of the available reconstructed samples used for prediction and therefore constitutes a subset of available reference samples. One would be motivated to combine Blanch’s feature extraction technique to be effective in improving the efficiency of chroma intra-prediction (Blanch col. 3, lines 15-16– to be effective in improving the efficiency of chroma intra-prediction).
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Lin, in view of Park, and further in view of Parkhomenko et al, US 20120155655 A1 (Parkhomenko).
Regarding Claim 9, Lin and Park teach the method of Claim 7, as outlined above.
However, Lin does not explicitly disclose wherein the extracting of the one or more features comprises: updating a histogram based on reference samples sequentially obtained while looping through the reference samples in the first loop, and recomputing, based on the updated histogram, the one or more features.
Parkhomenko teaches wherein the extracting of the one or more features comprises: updating a histogram based on reference samples sequentially obtained (Parkhomenko [0040]– In lines 7 to 25 of pseudocode 500 in FIG. 5A, the histogram is updated based on the samples Fn[i] of the received frame, where i=1, . . . , M) while looping through the reference samples in the first loop (Parkhomenko Fig. 5A– loop for 1 to size of frame in samples), and recomputing, based on the updated histogram, the one or more features (Parkhomenko [0043]– In lines 25 to 40 of pseudocode 500 in FIG. 5B, the...threshold ...is updated based on the generated histogram).
Therefore, it 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 to modify Lin to incorporate updating a histogram sequentially and recomputing features as taught by Parkhomenko.
One would be motivated to combine Parkhomenko’s updating of a histogram sequentially in order to achieve a recomputing of features (Parkhomenko [0040]– In lines 7 to 25 of pseudocode 500 in FIG. 5A, the histogram is updated based on the samples Fn[i] of the received frame, where i=1, . . . , M; Parkhomenko Fig. 5A– loop for 1 to size of frame in samples; Parkhomenko [0043]– In lines 25 to 40 of pseudocode 500 in FIG. 5B, the...threshold ...is updated based on the generated histogram).
Claim 10 and 12-14 are rejected under 35 U.S.C. 103 as being unpatentable over Lin, in view of Park, and further in view of Zhang et al, US 20200382769 A1 (Zhang).
Regarding Claim 10, Lin and Park teach the method of Claim 7, as outlined above.
However, Lin does not explicitly disclose wherein the one or more features include a threshold value computed based on the reference samples.
Zhang teaches wherein the one or more features include a threshold value computed based on the reference samples (Zhang ¶ [0055]– Threshold is calculated as the average value of the neighboring reconstructed luma samples).
Therefore, it 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 to modify Lin to incorporate a threshold value as taught by Zhang.
One would be motivated to combine Zhang’s threshold value to achieve a well-balanced trade-off between complexity and compression efficiency improvement (Zhang [0042]– approach that has a well-balanced trade-off between complexity and compression efficiency improvement).
Regarding Claim 12, Lin and Park teach the method of Claim 1, as outlined above.
However, Lin does not explicitly disclose wherein the models are one of a cross-component linear model (CCLM), a convolutional cross- component model (CCCM), a gradient and location based convolutional cross-component model (GL-CCCM), or a combination thereof.
Zhang teaches wherein the models are one of a cross-component linear model (CCLM), a convolutional cross- component model (CCCM), a gradient and location based convolutional cross-component model (GL-CCCM), or a combination thereof (Zhang ¶ [0043]– In some embodiments, and to reduce the cross-component redundancy, a cross-component linear model (CCLM) prediction mode (also referred to as LM), is used…).
Therefore, it 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 to modify Lin to incorporate a cross- component linear model as taught by Zhang.
One would be motivated to combine Zhang’s cross- component linear model to achieve a well-balanced trade-off between complexity and compression efficiency improvement. Please see the motivation of Claim 10.
Regarding Claim 13, Lin and Park teach the method of Claim 1, as outlined above.
However, Lin does not explicitly disclose wherein the method is performed by a video encoder.
Zhang teaches wherein the method is performed by a video encoder (Zhang ¶ [0856]– The examples described...may be implemented at a video encoder and/or decoder).
Therefore, it 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 to modify Lin to incorporate a video encoder as taught by Zhang.
One would be motivated to combine Zhang’s video encoder to achieve a well-balanced trade-off between complexity and compression efficiency improvement. Please see the motivation of Claim 10.
Regarding Claim 14, Lin and Park teach the method of Claim 1, as outlined above.
However, Lin does not explicitly disclose wherein the method is performed by a video decoder.
Zhang teaches wherein the method is performed by a video decoder (Zhang ¶ [0856]– The examples described...may be implemented at a video encoder and/or decoder).
Therefore, it 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 to modify Lin to incorporate a video decoder as taught by Zhang.
One would be motivated to combine Zhang’s video decoder to achieve a well-balanced trade-off between complexity and compression efficiency improvement. Please see the motivation of Claim 10.
Claim 11 is rejected under 35 U.S.C. 103 as being unpatentable over Lin, in view of Park, and further in view of Jhu et al, WO 2023239676 A1 (Jhu).
Regarding Claim 11, Lin and Park teach the method of Claim 1, as outlined above.
However, Lin does not explicitly disclose wherein the accumulated data of the first model include a first auto-correlation matrix and a first cross- correlation vector, and wherein the accumulated data of the second model include a second auto- correlation matrix and a second cross-correlation vector.
Jhu teaches wherein the accumulated data of the first model include a first auto-correlation matrix and a first cross- correlation vector (Jhu ¶ [00198]– The MSE minimization is performed by calculating autocorrelation matrix for the luma input and a cross-correlation vector between the luma input and chroma output), and wherein the accumulated data of the second model include a second auto- correlation matrix and a second cross-correlation vector (Jhu [00190]– Also, similarly to CCLM, there is an option of using...multi-model variant of CCCM. The multi -model variant uses two models…).
Therefore, it 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 to modify Lin to incorporate auto-correlation matrixes and a cross- correlation vectors as taught by Jhu.
Because the reference teaches a multi-model variant using two models, a person having ordinary skill in the art would understand that each model has its own corresponding auto-correlation matrix and cross-correlation vector. One would be motivated to combine Jhu’s auto-correlation matrixes and a cross- correlation vectors to improve coding efficiency (Jhu [0002]– improving the coding efficiency of the image/video blocks).
Claims 15, 17 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang, in view of Park.
Regarding Claim 15, Zhang discloses an apparatus, comprising: at least one processor (Zhang ¶ [0917]– An apparatus in a video system comprising a processor...); and memory storing instructions that, when executed by the at least one processor, cause the apparatus to (Zhang ¶ [0917]– An apparatus in a video system comprising a processor and a non-transitory memory with instructions thereon, wherein the instructions upon execution by the processor, cause the processor to implement the method): obtain video data (Zhang Fig. 21B ¶ [0865]– The system 3100 may include input 3102 for receiving video content), including data representing a video data region (Zhang ¶ [0977]– generating the coded representation from the current block), and compute models, used for cross-component based prediction of chroma samples from the video data region, the computing comprises (Zhang ¶ [0053]– CCLM mode employs one linear model for predicting the chroma samples from the luma samples for the whole CU, while in MMLM, there can be two models): accumulating data of a first model and of a second model through reference samples selected from the video data (Zhang ¶ [0054]– In MMLM, neighboring luma samples and neighboring chroma samples of the current block are classified into two groups, each group is used as a training set to derive a linear model).
However, Zhang does not explicitly disclose concurrently using a first loop.
Park teaches concurrently using a first loop (Park ¶ [0094]– Thus, two independent convolution operations are executed in parallel by the neural engines 314A and 314B in a single loop).
Therefore, it 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 to modify Zhang to incorporate accumulating data currently for two models in a loop as taught by Park.
One would be motivated to combine Park’s incorporation of accumulating data currently for two models to achieve data accumulation in a single loop. Please see motivation of Claim 1.
Regarding Claim 17, Zhang and Park teach the apparatus of Claim 15, as outlined above. In addition, Zhang discloses wherein the instructions further cause the system to: apply the first model and the second model through samples of the video region, wherein the first model and the second model are applied to predict the chroma samples (Zhang ¶ [0053]– CCLM mode employs one linear model for predicting the chroma samples from the luma samples for the whole CU, while in MMLM, there can be two models) according to their classification into a first class or a second class (Zhang ¶ [0054]– In MMLM, neighboring luma samples and neighboring chroma samples of the current block are classified into two groups, each group is used as a training set to derive a linear model).
However, Zhang does not explicitly disclose using a second loop.
Park teaches using a second loop (Park [0065]– The second...loop performs convolution operation for each slice in the input data).
Therefore, it 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 to modify Zhang to incorporate a second loop as taught by Park.
One would be motivated to combine Park’s incorporation of a second loop for further filtering. See motivation of Claim 3.
Regarding Claim 20, Zhang, in combination, further discloses the apparatus of Claim 15, wherein the accumulated data of the first model and of the second model (Zhang ¶ [0053]– CCLM mode employs one linear model for predicting the chroma samples from the luma samples for the whole CU, while in MMLM, there can be two models) are generated based on the reference samples according to their respective classification into a first class and a second class (Zhang ¶ [0054]– In MMLM, neighboring luma samples and neighboring chroma samples of the current block are classified into two groups, each group is used as a training set to derive a linear model).
Claim 16 is rejected under 35 U.S.C. 103 as being unpatentable over Zhang, in view Park, and further in view of Henderson.
Regarding Claim 16, Zhang and Park teach the apparatus of Claim 15, as outlined above. In addition, Zhang discloses of the first model and the second model (Zhang ¶ [0053]– CCLM mode employs one linear model for predicting the chroma samples from the luma samples for the whole CU, while in MMLM, there can be two models).
However, Zhang does not explicitly disclose wherein the instructions further cause the system to: rescale at least one parameter, wherein the rescaling includes rescaling the parameter into a lower bit size representation.
Henderson teaches wherein the instructions further cause the system to: rescale at least one parameter, wherein the rescaling includes rescaling the parameter into a lower bit size representation (Henderson [0237]– The first and second model are compressed using subword-level parameterisation and quantisation. Quantization of the stored embeddings, as well as optionally that of other neural network parameters, reduces model storage requirements).
Therefore, it 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 to modify Zhang to incorporate rescaling a parameter into a lower bit size representation as taught by Henderson.
Although Henderson does not expressly state a lower bit-size representation, a person of ordinary skill in the art would understand quantization to mean reducing the number of bits used to represent the parameters, thereby providing a lower bit-size representation. One would be motivated to combine Henderson’s incorporation of rescaling a parameter into a lower bit size representation to increase memory efficiency. Please see the motivation of Claim 2.
Claim 18 is rejected under 35 U.S.C. 103 as being unpatentable over Zhang, in view Park, and further in view of Pednault.
Regarding Claim 18, Zhang and Park teach the apparatus of Claim 15, as outlined above.
However, Zhang does not explicitly disclose wherein the instructions further cause the system to: combine the accumulated data of the first model and of the second model into a combined data of a single model.
Pednault teaches wherein the instructions further cause the system to: combine the accumulated data of the first model and of the second model into a combined data of a single model (Pednault ¶ [0429]– update the sufficient statistics stored in the model object by combining the existing sufficient statistics with those of the input model object, and to then construct a new model based on the updated sufficient statistics).
Therefore, it 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 to modify Zhang to incorporate combining data of two models into one as taught by Pednault.
One would be motivated to combine Pednault’s incorporation of combining data of two models into one to attain high levels of predictive accuracy. Please see the motivation of Claim 4.
Claim 19 is rejected under 35 U.S.C. 103 as being unpatentable over Zhang, in view Park, and further in view of Pednault and Henderson.
Regarding Claim 19, Zhang, Park and Pednault teach the apparatus of Claim 18, as outlined above.
However, Zhang in view of Park and Pednault does not explicitly disclose wherein the instructions further cause the system to: rescale at least one parameter, wherein the rescaling includes rescaling the parameter into a lower bit size representation.
Henderson teaches wherein the instructions further cause the system to: rescale at least one parameter, wherein the rescaling includes rescaling the parameter into a lower bit size representation (Henderson ¶ [0237]– The first and second model are compressed using subword-level parameterisation and quantisation. Quantization of the stored embeddings, as well as optionally that of other neural network parameters, reduces model storage requirements).
Therefore, it 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 to modify Zhang to incorporate rescaling a parameter into a lower bit size representation as taught by Henderson.
One would be motivated to combine Henderson’s incorporation of rescaling a parameter into a lower bit size representation to increase memory efficiency. Please see the motivation of Claim 2.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHRISTIAN P MCFALL whose telephone number is (571)270-0773. The examiner can normally be reached Monday Friday, 8 a.m. 5 p.m. ET..
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, Joseph G. Ustaris can be reached at (571) 272-7383. 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.
/C.P.M./Examiner, Art Unit 2483
/JOSEPH G USTARIS/Supervisory Patent Examiner, Art Unit 2483
Prosecution Insights