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 .
Abstract
Applicant may upload a wrong file.
The abstract should be in narrative form and generally limited to a single paragraph on a separate sheet within the range of 50 to 150 words in length.
Claim Objections
Claim 8 objected to because of the following informalities: “evaluation data and calculated data” in line 3. It appears that it should be “the evaluation data and the calculated data” Appropriate correction is required.
Claim 11 objected to because of the following informalities: “calculated data” in line 3. It appears that it should be “the calculated data” Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 8, 11, 16, 18-19 rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 8 recites the limitation "the pre-processing steps" in line 2. There is insufficient antecedent basis for this limitation in the claim because “pre-processing steps” is not defined or mentioned. Because the claim is indefinite and cannot be properly construed, for purposes of examination, “the method comprising” in claim 1 line 2 is being interpreted as " the method comprising pre-processing steps of". Appropriate clarification is required.
Claim 11 recites the limitation " the provided sample data" in line 3. There is insufficient antecedent basis for this limitation in the claim because “provided sample data” is not defined or mentioned. Because the claim is indefinite and cannot be properly construed, for purposes of examination, this limitation is being interpreted as " the generated sample data". Appropriate clarification is required.
Claim 16 recites the limitation “a machine learning model” in line 3. It is indefinite because it is not clear what relationship between the “a machine learning model” in line 3 and the “a machine learning model” in claim 12 line 2. Because the claim is indefinite and cannot be properly construed, for purposes of examination, this limitation is being interpreted as " the machine learning model ". Appropriate clarification is required.
Claim 18 recites the limitation “each feature” in line 1. It is indefinite because it is not clear whether or not the “each feature” in line 1 relates to the “any one or more features of the sample data” mentioned in claim 17 line 2. Because the claim is indefinite and cannot be properly construed, for purposes of examination, this limitation is being interpreted as " each feature of the one or more features ". Appropriate clarification is required.
Claim 19 recites the limitation “one feature” in line 1. It is indefinite because it is not clear whether or not the “one feature” in line 1 relates to the “any one or more features of the sample data” mentioned in claim 17 line 2 and/or the “each feature” in claim 18 line 1. Because the claim is indefinite and cannot be properly construed, for purposes of examination, this limitation is being interpreted as " one feature of the one or more features ". Appropriate clarification is required.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-19 are rejected under 35 U.S.C. 103 as being unpatentable over Chen et al. (US2022/0196798, hereafter Chen) in view of Roh (US 2021/0405150, hereafter Roh).
Regarding claim 1, Chen (‘798) discloses that A method for pre-processing data for further processing by a machine learning model { Fig.13 item 1302 input to item 1314 (NN detector); [0017] lines 1-2 (FIG . 13 illustrates a flowchart of an example method to perform object detection by the radar detector); [0221] lines 1-2 (The method 1300 , at block 1302 , may include generating a dataset)}, the method comprising:
generating sample data from a transmitted signal and a received signal over a time period { Fig.4 items 409 (mixer), 407 (chirps), 411 (ADC); [0133] line 2 (a sequence of chirps 407 ,); [0134] lines 3-5 (A mixer 409 of the radar frontend 401 mixes the radio transmit signal with the radio receive signal .)};
allocating the sample data to a plurality of range bins { Fig.5; [0140] line 7 (L range bins .)};
selecting a first subset of the range bins { Fig.5; [0139] lines 5-7 (For each chirp , the data cube 504 includes L samples ( e.g. L = 512 ) , which are arranged in the so - called " fast time ” -direction); [0140] line 7 (L range bins .); Examiner’s note: one chirp with L range bins for “a first subset of the range bins”};
generating evaluation data based on an evaluation of the sample data of the first subset of the range bins against one or more criteria { [0540] lines 1 (radar processing device 5100), 4-6 (determine which radar reception data values of the radar reception data values satisfy a predefined evaluation criterion); [0758] lines 7-9 (determine a measure according to a predefined evaluation metric of each radar reception data value); Examiner’s note: “each radar reception data value” for “the sample data of the first subset of the range bins”};
generating calculated data based on the sample data {Fig.13 item 1312 (radar pipeline simulation); Fig.33 item 3303 (R-D / AOA); [0226] lines 1-2 (The method 1300 , at block 1312 , may include simulating a radar pipeline.), 5-7 (the radar processor 1104 may simulate cross correlation , doppler estimation , beamforming , etc. as part of simulating the radar pipeline)}; and
providing the evaluation data and the calculated data to a machine learning model with temporal dynamics for further processing { Fig.13 item 1302 input to item 1314 (NN detector); Fig.31; Fig.32; Fig.33 item 3303 (R-D / AOA ), 3302 (neural network); [0035] lines 1-2 (Fig.31, a supervised training of a neural network); [0036] lines 1-2 (Fig.32, a self - supervised training of a neural network); Examiner’s note: training for “with temporal dynamics for further processing)}.
However, Chen (‘798) does not explicitly disclose (see words with underline) “selecting a second subset of range bins based on the evaluation data”, and “generating calculated data based on the sample data of the second subset of range bins ”. In the same field of endeavor, Roh (‘150) discloses that
selecting a second subset of range bins based on the evaluation data { [0037] lines 5-8 (values in the range profile vector that are peaks , that is , rise above the surrounding elements in the array ( e.g. , vector or matrix ) as well as a background level , can be selected as detected range bins .); Examiner’s note: “detected range bins” for “a second subset of range bins” }; and
generating calculated data based on the sample data of the second subset of range bins {[0036] lines 4-5 (To determine good candidates for DoA spectral estimation ,); [0037] lines 9-10 (DoA spectral estimation described previously can be performed only on the detected range bins)};
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Chen (‘798) with the teachings of Roh (‘150) {generate estimation (e.g. DoA spectral estimation ) by only processing data with object (e.g. peaks as detected range bins)} to generate estimation (e.g. DoA spectral estimation ) by only processing data with object (e.g. peaks as detected range bins). Doing so would reduce computation cost for DoA spectral estimation by running DoA estimation only on range bins that have high probability of having one or more objects, as recognized by Roh (‘150) {[0036] lines 1-4 (The computation cost for DoA spectral estimation can be reduced by running DoA estimation only on range bins that have high probability of having one or more objects)}.
Regarding claim 2, which depends on claim 1, the combination of Chen (‘798) and Roh (‘150) discloses that in the method,
the first subset of range bins for the time period is selected on the basis of sample data generated during a previous time period {see Chen (‘798) Fig.5 slow time, fast time; [0139] lines 7-9 (The data cube 504 includes samples for K chirps , which are arranged in the so – called “ slow time ” -direction .); Examiner’s note: Fig.5 shows that data cube has same row and column, therefore each chirp has same range bins with other chirps}.
Regarding claim 3, which depends on claim 1, Chen (‘798) does not explicitly disclose “the criteria used to determine the second subset of range bins comprises whether the sample data of a respective range bin indicates an object of interest has been detected”. In the same field of endeavor, Roh (‘150) discloses that in the method,
the criteria used to determine the second subset of range bins comprises whether the sample data of a respective range bin indicates an object of interest has been detected { [0037] lines 1-8 (the range profile vector is determined , a one - dimensional CFAR algorithm can be performed on the range profile vector s to determine which range bins are likely to contain objects, referred to herein as “detected range bins”. values in the range profile vector that are peaks , that is , rise above the surrounding elements in the array ( e.g. , vector or matrix ) as well as a background level , can be selected as detected range bins .}.
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Chen (‘798) with the teachings of Roh (‘150) {generate estimation (e.g. DoA spectral estimation ) by only processing data with object (e.g. peaks as detected range bins) using criteria} to generate estimation (e.g. DoA spectral estimation ) by only processing data with object (e.g. peaks as detected range bins) using criteria. Doing so would reduce computation cost for DoA spectral estimation by running DoA estimation only on range bins that have high probability of having one or more objects, as recognized by Roh (‘150) {[0036] lines 1-4 (The computation cost for DoA spectral estimation can be reduced by running DoA estimation only on range bins that have high probability of having one or more objects)}.
Regarding claim 4, which depends on claim 1, Chen (‘798) does not explicitly disclose “the evaluation of sample data of the first subset of range bins comprises comparing a magnitude of the sample data of a range bin with a threshold”. In the same field of endeavor, Roh (‘150) discloses that in the method,
the evaluation of sample data of the first subset of range bins comprises comparing a magnitude of the sample data of a range bin with a threshold { [0037] lines 5-8 (values in the range profile vector that are peaks , that is , rise above the surrounding elements in the array ( e.g. , vector or matrix ) as well as a background level , can be selected as detected range bins .)}.
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Chen (‘798) with the teachings of Roh (‘150) {generate estimation (e.g. DoA spectral estimation ) by only processing data with object (e.g. peaks as detected range bins) using magnitude (e.g. peak) compared with a threshold (e.g. a background level)} to generate estimation (e.g. DoA spectral estimation ) by only processing data with object (e.g. peaks as detected range bins) using magnitude (e.g. peak) compared with a threshold (e.g. a background level). Doing so would reduce computation cost for DoA spectral estimation by running DoA estimation only on range bins that have high probability of having one or more objects, as recognized by Roh (‘150) {[0036] lines 1-4 (The computation cost for DoA spectral estimation can be reduced by running DoA estimation only on range bins that have high probability of having one or more objects)}.
Regarding claim 5, which depends on claims 1 and 4, Chen (‘798) does not explicitly disclose “the threshold is adjusted based on a computed probability of false detection”. In the same field of endeavor, Roh (‘150) discloses that in the method,
the threshold is adjusted based on a computed probability of false detection { [0037] lines 1-8 (the range profile vector is determined , a one - dimensional CFAR algorithm can be performed on the range profile vector s to determine which range bins are likely to contain objects, referred to herein as “detected range bins”. values in the range profile vector that are peaks , that is , rise above the surrounding elements in the array ( e.g. , vector or matrix ) as well as a background level , can be selected as detected range bins .}.
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Chen (‘798) with the teachings of Roh (‘150) {generate estimation (e.g. DoA spectral estimation ) by only processing data with object (e.g. peaks as detected range bins) using magnitude (e.g. peak) compared with an adjusted threshold (e.g. a background level)} to generate estimation (e.g. DoA spectral estimation ) by only processing data with object (e.g. peaks as detected range bins) using magnitude (e.g. peak) compared with an adjusted threshold (e.g. a background level). Doing so would reduce computation cost for DoA spectral estimation by running DoA estimation only on range bins that have high probability of having one or more objects via a constant false alarm rate ( CFAR ) algorithm, as recognized by Roh (‘150) {[0036] lines 1-6 (The computation cost for DoA spectral estimation can be reduced by running DoA estimation only on range bins that have high probability of having one or more objects, a one - dimensional constant false alarm rate ( CFAR ) algorithm can be used)}.
Regarding claim 6, which depends on claim 1, the combination of Chen (‘798) and Roh (‘150) discloses that in the method,
the calculated data is angle of arrival data {see Chen (‘798) Fig.33 item 3303 (R-D / AOA map); [0145] lines 3-4 (the radar processor 309 further to determine the angle of arrival); [0408] lines 2-3 (a first range Doppler map 3303); [0409] line 3 (from a first angle of arrival map 3303)}.
Regarding claim 7, which depends on claims 1 and 6, the combination of Chen (‘798) and Roh (‘150) discloses that in the method,
the angle of arrival data is calculated on the basis of first sample data generated from a first received signal from a first receive antenna, and second sample data generated from a second received signal from a second receive antenna {see Chen (‘798) Fig.6; Fig.7; Fig.11 item 1108 (Rx Ant.1, Rx Ant. N), 1117 (Radar Pipeline); [0010] lines 1-3 (Fig.6, the determination of an angle of arrival of an incoming radio signal received by a receive antenna array ;); [0011] lines 1-3 (Fig.7, a radar MIMO ( multiple – input multiple - output ) arrangement including a transmit antenna array and a receive antenna array) }.
Regarding claim 8, which depends on claim 1, the combination of Chen (‘798) and Roh (‘150) discloses that the method further comprising
repeating the pre-processing steps for a series of consecutive time periods to provide a series of evaluation data and calculated data associated with the series of consecutive time periods to the machine learning model for further processing {see Chen (‘798) Fig.5 item 504 (data cube); [0036] lines 4-5 (To determine good candidates for DoA spectral estimation ,); [0037] lines 9-10 (DoA spectral estimation described previously can be performed only on the detected range bins); [0133] line 2 (a sequence of chirps 407 ,); [0139] lines 7-9 (The data cube 504 includes samples for K chirps , which are arranged in the so – called “ slow time ” -direction .); [0336] lines 5-7 (the radar measurement data may include , for each frame , radar measurement data for a predetermined number of chirps ( e.g. K chirps as in the example of FIG . 5 ) .); [0352] line 5 (a sequence of radar measurement frames); [0540] lines 1 (radar processing device 5100), 4-6 (determine which radar reception data values of the radar reception data values satisfy a predefined evaluation criterion); [0758] lines 7-9 (determine a measure according to a predefined evaluation metric of each radar reception data value); Examiner’s note: “each radar reception data value” for “the sample data of the first subset of the range bins”}.
Regarding claim 9, which depends on claim 1, the combination of Chen (‘798) and Roh (‘150) discloses that in the method,
the transmitted signal and received signal are radar signals {see Chen (‘798) Fig.4 items 402 (radar processor), 406 (transmit antenna); Fig.11 items 1104 (radar processor), 1108 (Tx, Rx); [0008] lines 1-2 (Fig.4, FMCW, radar device); [0132] line 3 (transmit antenna 406); [0134] lines 1-2 (receive antennas 408)}.
Regarding claim 10, which depends on claim 1, the combination of Chen (‘798) and Roh (‘150) discloses that in the method,
the transmitted signal is a frequency modulated continuous wave radar signal { see Chen (‘798) Fig.4 items 406 (transmit antenna), 407 (); [0008] lines 1-2 (Fig.4, FMCW, radar device); [0132] line 3 (transmit antenna 406); [0133 line 2 (a sequence of chirps 407 ,)]}, and
the time period is determined based on the period of a chirp of the transmitted signal { see Chen (‘798) Fig.4 item 403}.
Regarding claim 11, which depends on claim 1, the combination of Chen (‘798) and Roh (‘150) discloses that in the method,
the machine learning model is adapted to perform object or gesture detection or recognition based on the provided sample data and calculated data for a sequence of time periods {see Chen (‘798) Fig.13 items 1302, 1314 (NN detector); Fig.33 item 3303 (R-D / AOA ); [0221] lines 1-2 (The method 1300 , at block 1302 , may include generating a dataset); [0227] lines 1-3 (The method 1300 , at block 1314 , may include performing object detection using a neural network ( e.g. , NN ) .); [0348] lines 11-14 (NN models that provides functionalities such as activity detection , gesture recognition , high fidelity object detection along with tracking information of moving objects specifically slow moving ones)}.
Regarding claim 12, as modified above, Chen (‘798) discloses that A system {Fig.4; Fig.13 item 1302; [0071] lines 8-9 (a computing system); [0082] lines 1-2 (radar system)} for pre-processing data for further processing by a machine learning model with temporal dynamics { Fig.13 item 1302 input to item 1314 (NN detector); Fig.31; Fig.32; Fig.33 item 3303 (R-D / AOA ); [0017] lines 1-2 (FIG . 13 illustrates a flowchart of an example method to perform object detection by the radar detector); [0035] lines 1-2 (Fig.31, a supervised training of a neural network); [0036] lines 1-2 (Fig.32, a self - supervised training of a neural network); Examiner’s note: “training” for “with temporal dynamics)}, the system being arranged to receive a transmitted signal and a received signal {Fig.4; [0082] lines 1-2 (radar system)}, and the system comprising a processor {Fig.4 item 402 (radar processor)} configured to:
generate sample data from a transmitted signal and a received signal over a time period;
allocate the sample data to a plurality of range bins;
select a first subset of the range bins;
generate evaluation data based on an evaluation of the sample data of the first subset of the range bins against one or more criteria;
select a second subset of range bins based on the evaluation data; and
generate calculated data based on the sample data of the second subset of range bins; and the system being configured to provide the evaluation data and the calculated data to a machine learning model with temporal dynamics for further processing.
{The claim limitations above are the same or substantially the same scope as the corresponding claim limitations in claim 1. Therefore the claim limitations above are rejected in the same or substantially the same manner as in claim 1. See the rejections of claim 1}.
Regarding claims 13-14, Applicant recites claim limitations of the same or substantially the same scope as that of claims 2-3, respectively. Accordingly, claims 13-14 are rejected in the same or substantially the same manner as claims 2-3, respectively, shown above.
Regarding claim 15, Applicant recites claim limitations of the same or substantially the same scope as that of claim 8. Accordingly, claim 15 is rejected in the same or substantially the same manner as claim 8, shown above.
Regarding claim 16, which depends on claim 12, the combination of Chen (‘798) and Roh (‘150) discloses that the system further comprising
a sensor including one or more transmit antennas and one or more receive antennas {see Chen (‘798) Fig.11 items 1140 (radar processor), 1108 (Tx Ant. 1, Tx Ant. M, Rx Ant. 1, Rx Ant. N)}, and
a machine learning model, integrated in a single semiconductor chip {see Chen (‘798) [0872] lines 1-7 (the above descriptions and connected figures may depict electronic device components as separate elements , combine or integrate discrete elements into a single element . Such may include combining two or more circuits for form a single circuit , mounting two or more circuits onto a common chip)}.
Regarding claim 17, which depends on claim 1, the combination of Chen (‘798) and Roh (‘150) discloses that in the method,
the criteria used for the evaluation are based on any one or more features of the sample data {see Chen (‘798) [0838] lines 4-6 (radar reception data values of the radar reception data values satisfy a predefined evaluation criterion ); [0839] lines 2-3 (the predefined evaluation criterion is a predefined power criterion); Examiner’s note: “data values” and “power” for “any one or more features of the sample data”}.
Regarding claim 18, which depends on claims 1 and 17, Chen (‘798) does not explicitly disclose “each feature is calculated over a separate selection of the sample data”. In the same field of endeavor, Roh (‘150) discloses that in the method,
each feature is calculated over a separate selection of the sample data { [0037] lines 1-8 (the range profile vector is determined , a one - dimensional CFAR algorithm can be performed on the range profile vector s to determine which range bins are likely to contain objects, referred to herein as “detected range bins”. values in the range profile vector that are peaks , that is , rise above the surrounding elements in the array ( e.g. , vector or matrix ) as well as a background level , can be selected as detected range bins); Examiner’s note: CFAR algorithm for “a separate selection of the sample data” because the criteria used in CFAR algorithm is based on detected results.}.
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Chen (‘798) with the teachings of Roh (‘150) {generate estimation (e.g. DoA spectral estimation ) by only processing data with object (e.g. peaks as detected range bins) using magnitude (e.g. peak) compared with an adjusted threshold (e.g. a background level)} to generate estimation (e.g. DoA spectral estimation ) by only processing data with object (e.g. peaks as detected range bins) using magnitude (e.g. peak) compared with an adjusted threshold (e.g. a background level). Doing so would reduce computation cost for DoA spectral estimation by running DoA estimation only on range bins that have high probability of having one or more objects via a constant false alarm rate ( CFAR ) algorithm, as recognized by Roh (‘150) {[0036] lines 1-6 (The computation cost for DoA spectral estimation can be reduced by running DoA estimation only on range bins that have high probability of having one or more objects, a one - dimensional constant false alarm rate ( CFAR ) algorithm can be used)}.
Regarding claim 19, which depends on claims 1 and 17-18, Chen (‘798) does not explicitly disclose “one feature is calculated for the first subset of samples and another feature is calculated for the second set of samples, to allow fine-tuning of range bin sampling resolution and processing cost per feature”. In the same field of endeavor, Roh (‘150) discloses that in the method,
one feature is calculated for the first subset of samples and another feature is calculated for the second set of samples, to allow fine-tuning of range bin sampling resolution and processing cost per feature { [0012] lines 6-7 (clutter removal algorithm and allows high - resolution directional - of - arrival ( DOA ) estimation); [0036] lines 1-4 (The computation cost for DoA spectral estimation can be reduced by running DoA estimation only on range bins that have high probability of having one or more objects); [0037] lines 1-10 (the range profile vector is determined , a one - dimensional CFAR algorithm can be performed on the range profile vector s to determine which range bins are likely to contain objects, referred to herein as “detected range bins”. values in the range profile vector that are peaks , that is , rise above the surrounding elements in the array ( e.g. , vector or matrix ) as well as a background level , can be selected as detected range bins . The DoA spectral estimation described previously can be performed only on the detected range bins); Examiner’s note: “background level” for “one feature is calculated for the first subset of samples”. “peaks” for “another feature is calculated for the second set of samples”}.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to YONGHONG LI whose telephone number is (571)272-5946. The examiner can normally be reached 8:30am - 5:00pm.
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, Vladimir Magloire can be reached at (571)270-5144. 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.
/YONGHONG LI/ Examiner, Art Unit 3648
Prosecution Insights