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 .
Response to Amendment
The amendments filed June 9, 2026 have been entered. Claims 1-3, 5-15, and 17-26 are currently pending in this application. Claims 1, 13, and 24 have been amended. Claims 4 and 16 have been cancelled. Claims 25-26 are new.
Response to Arguments
Applicant’s arguments, see pages 7-8, filed June 9, 2026, with respect to the rejections of claims 1, 13, and 24 under 35 U.S.C. 103 have been fully considered and are persuasive. Therefore, the rejections have been withdrawn. However, upon further consideration, new grounds of rejection are made in view of Di et al. (US 20180203106 A1).
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 1-2, 7, 10-14, 19, and 22-25 are rejected under 35 U.S.C. 103 as being unpatentable over Di et al. (US 20180203106 A1), hereinafter Di, in view of Aydogdu et al. (Multi-Modal Cross Learning for Improved People Counting using Short-Range FMCW Radar [2020]), hereinafter Aydogdu..
Regarding claim 1, Di teaches an apparatus, comprising:
a processor, and a memory coupled to the processor with instructions stored thereon (see para. 18), wherein the instructions, when executed by the processor, enable the apparatus to:
obtain radar data indicating a received radar signal of a radar sensor, obtain data indicating a detection zone in which persons are to be detected, modify the radar data by masking an undesired zone outside the detection zone, masking comprising at least one of attenuating, removing, or zeroing a part of the radar data corresponding to the undesired zone (paras. 28-30, “In step 312, data is received from the at least one detection device, the detection device being configured to detect objects located behind the vehicle. As previously mentioned, the detection device may be one or more radar systems. In step 314, the method 310 performs pre-processing on the data from the at least one detection device. This pre-processing may include filtering the data to remove targets outside a region of interest, which will be described in greater detail in FIG. 4. […] In step 318, the targets are used to identify if a cluster exists. […] These clusters generally include targets that are within a specified distance from each other. In addition, these clusters are formed only when the number of targets within a cluster exceeds a threshold value.”; see para. 2 for evidence that the targets may be people), but fails to teach
input the modified radar data into a trained neural network, and
determine, using the trained neural network, a number of persons within the detection zone based on the modified radar data.
However, Aydogdu teaches wherein the instructions, when executed by the processor, enable the apparatus to:
input the radar data into a trained neural network, determine, using a trained neural network, a number of persons within the detection zone based on the radar data (page 250, “Radar systems enable remote-less sensing of multiple persons in its field of view. In this paper, we propose a novel people counting system using 60-GHz frequency modulated continuous wave radar sensor. The proposed deep convolutional neural network learns from supervised radar data and also through knowledge distillation via multi-modal cross-learning of representation from a synchronized camera-based deep convolutional neural network.”; page 251, “In this paper, we propose a novel multi-modal cross-learning framework for people counting application using frequency modulated continuous wave [FMCW] radar, which is trained not only from supervised radar data but also learns its parameters through high-level features distilled from camera-based DCNN. We demonstrate the people counting performance of our proposed solution with up to 4 people counting and detection of more than 4 people in indoor environment, which surpasses the performance achievable through its counterpart radar DCNN exploiting supervised radar-alone data.”), where Di teaches the modification of said radar data.
Di and Aydogdu are considered to be analogous to the claimed invention because they are in the same field of radar-based occupancy determination. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Di with the teachings of Aydogdu with the motivation of achieving high-accuracy occupancy determination.
Regarding claims 2, 14 and 25, Di in view of Aydogdu teaches the apparatus of claim 1, the method of claim 13, and the apparatus of claim 24 respectively,
wherein the data indicates at least one of a range interval or an angle interval in which persons are to be detected (Di; para. 38, “As such, a region of interest 132 is located behind the vehicle 112. Here, the region of interest is defined by an origin 134 that is defined by the x coordinates 136 and the y coordinates 138. The region of interest may extend in the y-direction 138 by a certain distance. In this example, the distance is 7 meters. In the x direction 136, the distance may be 0.5 meters in one direction from the origin 134 and 2.9 meters the other direction from the origin 134.”).
Regarding claims 7 and 19, Di in view of Aydogdu teaches the apparatus of claim 1 and the method of claim 13 respectively, but Di fails to teach wherein the instructions, when executed by the processor, further enable the apparatus to:
determine macro motion data indicating a macro motion in a field of view of the radar sensor based on the radar data and micro motion data indicating a micro motion in the field of view of the radar sensor based on the radar data,
determine a respective range-angle representation for the macro motion data and the micro motion data, and
modify the radar data by modifying the respective range-angle representation for the macro motion data and the micro motion data.
However, Aydogdu teaches wherein the instructions, when executed by the processor, further enable the apparatus to:
determine macro motion data indicating a macro motion in a field of view of the radar sensor based on the radar data and micro motion data indicating a micro motion in the field of view of the radar sensor based on the radar data, determine a respective range-angle representation for the macro motion data and the micro motion data, and modify the radar data by modifying the respective range-angle representation for the macro motion data and the micro motion data (page 250, “Radar systems enable remote-less sensing of multiple persons in its field of view. In this paper, we propose a novel people counting system using 60-GHz frequency modulated continuous wave radar sensor. The proposed deep convolutional neural network learns from supervised radar data and also through knowledge distillation via multi-modal cross-learning of representation from a synchronized camera-based deep convolutional neural network.”; Fig. 3, pre-processing pipeline to generate separate range-angle image using micro-Doppler and macro-Doppler components and is fed as separate channels to the deep neural net; Fig. 4, range-angle images generated from macro-Doppler and micro-Doppler components respectively).
Di and Aydogdu are considered to be analogous to the claimed invention because they are in the same field of radar-based occupancy determination. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Di with the teachings of Aydogdu with the motivation of being able to output both rougher, large-scale motion and finer, small-scale motion data.
Regarding claims 10 and 22, Di in view of Aydogdu teaches the apparatus of claim 1 and the method of claim 13 respectively, but Di fails to teach wherein the instructions, when executed by the processor, further enable the apparatus to:
filter the radar data using a moving target indicator, and
modify the radar data by modifying the filtered radar data.
However, Aydogdu teaches wherein the instructions, when executed by the processor, further enable the apparatus to:
filter the radar data using a moving target indicator, and modify the radar data by modifying the filtered radar data (pages. 250-251, “The intermediate frequency signal from a chirp with NTS=128 number of samples and PN=64 consecutive chirps are collected and arranged in the form of a 2D matrix, as PN×NTS. As a first step, a range-Doppler image [RDI] is generated by subtracting the mean along fast time, followed by 1D Fast Fourier Transform [FFT] along fast time for all the PN chirps to obtain the range transformations. Following which mean across slow-time is subtracted followed by 1D FFT to obtain the Doppler transformation for all range bins. The RDI is then processed through moving target indicator [MTI] filter to removes reflections from any static targets, such as chairs and furnitures in the room. Once the RDI across both received channel NRx=2 is computed, the range-angle image [RAI] is computed through digital beam-forming algorithm utilizing the derived weights from angle model as follows: […] The first summation in eq. (1) transforms the RDI across each virtual channel into RDI across the angle space, and the second summation marginalizes across Doppler bins to generate the RAI.”; output of MTI filter is modified/processed to generated range-angle image).
Di and Aydogdu are considered to be analogous to the claimed invention because they are in the same field of radar-based occupancy determination. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Di with the teachings of Aydogdu with the motivation of being able to distinguish desired motion data.
Regarding claims 11, Di in view of Aydogdu teaches a radar system, comprising:
the apparatus according to claim 1 (Di; see rejection of claim 1 above), and
the radar sensor, wherein the radar sensor is configured to generate the radar data based on the received radar signal (Di; para. 17, “The detection devices 122A and/or 122B may be radar devices that send out radar signals. Any objects receiving these radar signals generally bounce these signals back to the detection devices 122A and/or 122B. This returned signal, when properly processed, can be utilized to determine the presence of an object or objects.”).
Regarding claim 12, Baker in view of Aydogdu teaches an electronic device, comprising:
the radar system according to claim 11 (Di; see rejection of claim 11 above), and
control circuitry configured to control the electronic device based on the determined number of persons (Di; para. 35, “In step 322, a determination is made when the trailer is located behind the vehicle based on the cluster features, the vehicle state, and/or the global features. The global features may be used for enhancement proposes. This determination may be made by setting a threshold [confidence level] for the global features and the cluster features, wherein exceeding the threshold is indicated that the trailer is located behind the vehicle.”; a determination is made depending on clustering which depends on a number of detected targets within a region within the ROI; see para. 45 for evidence of control circuitry).
Regarding claim 13, Di teaches a method, comprising:
obtaining radar data indicating a received radar signal of a radar sensor, obtaining data indicating a detection zone in which persons are to be detected, modifying the radar data by masking an undesired zone outside the detection zone, masking comprising at least one of attenuating, removing, or zeroing a part of the radar data corresponding to the undesired zone (paras. 28-30, “In step 312, data is received from the at least one detection device, the detection device being configured to detect objects located behind the vehicle. As previously mentioned, the detection device may be one or more radar systems. In step 314, the method 310 performs pre-processing on the data from the at least one detection device. This pre-processing may include filtering the data to remove targets outside a region of interest, which will be described in greater detail in FIG. 4. […] In step 318, the targets are used to identify if a cluster exists. […] These clusters generally include targets that are within a specified distance from each other. In addition, these clusters are formed only when the number of targets within a cluster exceeds a threshold value.”; see para. 2 for evidence that the targets may be people), but fails to teach
inputting the modified radar data into a trained neural network, and
determining, using a trained neural network, a number of persons within the detection zone based on the modified radar data.
However, Aydogdu teaches
inputting the radar data into a trained neural network, and determining, using a trained neural network, a number of persons within the detection zone based on the radar data (page 250, “Radar systems enable remote-less sensing of multiple persons in its field of view. In this paper, we propose a novel people counting system using 60-GHz frequency modulated continuous wave radar sensor. The proposed deep convolutional neural network learns from supervised radar data and also through knowledge distillation via multi-modal cross-learning of representation from a synchronized camera-based deep convolutional neural network.”; page 251, “In this paper, we propose a novel multi-modal cross-learning framework for people counting application using frequency modulated continuous wave [FMCW] radar, which is trained not only from supervised radar data but also learns its parameters through high-level features distilled from camera-based DCNN. We demonstrate the people counting performance of our proposed solution with up to 4 people counting and detection of more than 4 people in indoor environment, which surpasses the performance achievable through its counterpart radar DCNN exploiting supervised radar-alone data.”), where Di teaches the modification of said radar data.
Di and Aydogdu are considered to be analogous to the claimed invention because they are in the same field of radar-based occupancy determination. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Di with the teachings of Aydogdu with the motivation of achieving high-accuracy occupancy determination.
Regarding claim 23, Di in view of Aydogdu teaches
a non-transitory machine-readable medium having instructions stored thereon, wherein the instructions, when executed by a processor or programmable hardware, enable the processor or programmable hardware to perform the method of claim 13 (Di; para. 18, “The memory 130 may be any memory device capable of storing digital information.”).
Regarding claim 24, Di teaches an apparatus, comprising processing circuitry (see para. 18) configured to:
obtain radar data indicating a received radar signal of a radar sensor, obtain data indicating a detection zone in which persons are to be detected, modify the radar data by masking an undesired zone outside the detection zone, masking comprising at least one of attenuating, removing, or zeroing a part of the radar data corresponding to the undesired zone (paras. 28-30, “In step 312, data is received from the at least one detection device, the detection device being configured to detect objects located behind the vehicle. As previously mentioned, the detection device may be one or more radar systems. In step 314, the method 310 performs pre-processing on the data from the at least one detection device. This pre-processing may include filtering the data to remove targets outside a region of interest, which will be described in greater detail in FIG. 4. […] In step 318, the targets are used to identify if a cluster exists. […] These clusters generally include targets that are within a specified distance from each other. In addition, these clusters are formed only when the number of targets within a cluster exceeds a threshold value.”; see para. 2 for evidence that the targets may be people), but fails to teach processing circuitry configured to:
input the modified radar data into a trained neural network, and
determine, using the trained neural network, a number of persons within the detection zone based on the modified radar data.
However, Aydogdu teaches processing circuitry configured to:
input the modified radar data into a trained neural network, and
determine, using the trained neural network, a number of persons within the detection zone based on the modified radar data (page 250, “Radar systems enable remote-less sensing of multiple persons in its field of view. In this paper, we propose a novel people counting system using 60-GHz frequency modulated continuous wave radar sensor. The proposed deep convolutional neural network learns from supervised radar data and also through knowledge distillation via multi-modal cross-learning of representation from a synchronized camera-based deep convolutional neural network.”; page 251, “In this paper, we propose a novel multi-modal cross-learning framework for people counting application using frequency modulated continuous wave [FMCW] radar, which is trained not only from supervised radar data but also learns its parameters through high-level features distilled from camera-based DCNN. We demonstrate the people counting performance of our proposed solution with up to 4 people counting and detection of more than 4 people in indoor environment, which surpasses the performance achievable through its counterpart radar DCNN exploiting supervised radar-alone data.”), where Di teaches the modification of said radar data.
Di and Aydogdu are considered to be analogous to the claimed invention because they are in the same field of radar-based occupancy determination. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Di with the teachings of Aydogdu with the motivation of achieving high-accuracy occupancy determination.
Claims 3, 15, and 26 are rejected under 35 U.S.C. 103 as being unpatentable over Di in view of Aydogdu and further in view of Nishiyama (US 20120127013 A1).
Regarding claims 3, 15, and 26, Di in view of Aydogdu teaches the apparatus of claim 1, the method of claim 13, and the apparatus of claim 24 respectively, but fails to teach wherein the instructions, when executed by the processor, further enable the apparatus to
modify the radar data to exhibit a predefined data size.
However, Nishiyama teaches wherein the instructions, when executed by the processor, further enable the apparatus to
modify the radar data to exhibit a predefined data size (Abstract, “This disclosure provides a radar device, which detects an object of interest from image data produced based on an input signal containing echo signals caused by transmitted signals reflecting on objects.”; para. 13, “In one embodiment, when the object of interest is a flying object smaller than a predetermined size, the template image data may be produced so that the brightness of the image data is below a threshold and the particle size of the image data is above a threshold.”; para. 99, “FIG. 10 is a view showing an image resulting of the labeling process of a bird image data. Note that, although the image shown in FIG. 10 is extracted based on the bird template image data 14B, it also contains target echoes other than birds, for example, buoys, bank, or noise, which are similar to the template image data 14B. Based on the labeling results, the crowd rate calculation module 26 determines representative points [black dots in the drawing] in the classified groups, counts the number of the representative points contained in a fixed-sized area [line blocks in the drawing], and sets the count result as the crowd rate of the birds.”).
Di, Aydogdu, and Nishiyama are considered to be analogous to the claimed invention because they are in the same field of radar-based occupancy determination. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Di in view of Aydogdu with the teachings of Nishiyama with the motivation of achieving data of a predefined uniform size.
Claims 5-6, 8, 17-18, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Di in view of Aydogdu and further in view of Baker et al. (US 20230126895 A1), hereinafter Baker.
Regarding claims 5 and 17, Di in view of Aydogdu teaches the apparatus of claim 1 and the method of claim 13 respectively, but fails to teach
wherein the instructions, when executed by the processor, further enable the apparatus to obtain the data by receiving a user input of a user of the apparatus.
However, Baker teaches
wherein the instructions, when executed by the processor, further enable the apparatus to obtain the data by receiving a user input of a user of the apparatus (Baker; para. 84, “The occupant detection sensor 400 may comprise a user interface 424 including one or more actuators that may be used to configure the occupant detection sensor [e.g., during the commissioning procedure of the load control system 100 of FIG. 1]. For example, the user interface 424 may comprise one or more configuration buttons configured to be actuated to cycle through options that define the region of interest of the occupant detection sensor 400. In addition, the user interface 424 may comprise a potentiometer having a knob and/or a digital rotary switch configured to be rotated to adjust a value that defines the region of interest of the occupant detection sensor 400 [e.g., such as the rotation angle θR. Further, the user interface 424 may comprise other input devices, such as a digital DIP switch.”).
Di, Aydogdu, and Baker are considered to be analogous to the claimed invention because they are in the same field of radar-based occupancy determination. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Di in view of Aydogdu with the teachings of Baker with the motivation of allowing user control over the detection process.
Regarding claims 6 and 18, Di in view of Aydogdu teaches the apparatus of claim 1 and the method of claim 13 respectively, but fails to teach wherein the instructions, when executed by the processor, further enable the apparatus to:
determine a range-angle representation of the radar data, and modify the radar data by modifying the range-angle representation.
However, Baker teaches wherein the instructions, when executed by the processor, further enable the apparatus to:
determine a range-angle representation of the radar data, and modify the radar data by modifying the range-angle representation (para. 9, “In some examples, the occupant map may include a two-dimensional [2D] radar image indicating the locations of the occupants within a coverage area of the occupant detection circuit. The control circuit may be configured to generate the occupant count for the space based on feedback from the occupant detection circuit. The control circuit may be configured to maintain the occupant count in a region of interest of the based on whether the locations of the occupants are within the region of interest or not. The masked regions may be located within the region of interest but detected occupants within the masked region are excluded from the occupant count. The control circuit may be configured to report the occupant count to a system controller.”; para. 43, “The occupant detection circuit may be configured to determine the locations of an occupant as coordinates in a two-dimensional or three-dimensional coordinate system, e.g., a Cartesian or polar coordinate system [e.g., a 2D Cartesian coordinate system, 3D Cartesian coordinate system (e.g., with X, Y, and Z coordinates), a 2D polar coordinate system, or a 3D polar coordinate system (e.g., a spherical coordinate system.)].”).
Di, Aydogdu, and Baker are considered to be analogous to the claimed invention because they are in the same field of radar-based occupancy determination. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Di in view of Aydogdu with the teachings of Baker with the motivation of being able to resolve closely spaced targets.
Regarding claims 8 and 20, Di in view of Aydogdu and further in view of Baker teaches the apparatus of claim 6 and the method of claim 18 respectively, but Di fails to teach wherein the instructions, when executed by the processor, further enable the apparatus to:
determine the range-angle representation by determining a range-Doppler representation of the radar data, and perform channel processing on the range-Doppler representation.
However, Aydogdu teaches wherein the instructions, when executed by the processor, further enable the apparatus to:
determine the range-angle representation by determining a range-Doppler representation of the radar data, and perform channel processing on the range-Doppler representation (pages. 250-251, “The intermediate frequency signal from a chirp with NTS=128 number of samples and PN=64 consecutive chirps are collected and arranged in the form of a 2D matrix, as PN×NTS. As a first step, a range-Doppler image [RDI] is generated by subtracting the mean along fast time, followed by 1D Fast Fourier Transform [FFT] along fast time for all the PN chirps to obtain the range transformations. Following which mean across slow-time is subtracted followed by 1D FFT to obtain the Doppler transformation for all range bins. The RDI is then processed through moving target indicator [MTI] filter to removes reflections from any static targets, such as chairs and furnitures in the room. Once the RDI across both received channel NRx=2 is computed, the range-angle image [RAI] is computed through digital beam-forming algorithm utilizing the derived weights from angle model as follows: […] The first summation in eq. (1) transforms the RDI across each virtual channel into RDI across the angle space, and the second summation marginalizes across Doppler bins to generate the RAI.”).
Di, Aydogdu, and Baker are considered to be analogous to the claimed invention because they are in the same field of radar-based occupancy determination. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Di in view of Baker with the teachings of Aydogdu with the motivation of being able to distinguish targets in processing radar motion data.
Claims 9 and 21 are rejected under 35 U.S.C. 103 as being unpatentable over Di in view of Aydogdu and further in view of Baker and Boles (US 4546355 A).
Regarding claims 9 and 21, Di in view of Aydogdu and further in view of Baker teaches the apparatus of claim 6 and the method of claim 18 respectively, but fails to teach
wherein the instructions, when executed by the processor, further enable the apparatus to determine the range-angle representation using coherent integration.
However, Boles teaches
wherein the instructions, when executed by the processor, further enable the apparatus to determine the range-angle representation using coherent integration (col. 10 line 61 – col. 11 line 1, “Accordingly, on odd pulse repetition intervals [pri's], the signals from antenna arrays 1 and 2 are coherently added in a microwave hybrid summing network located in Azimuth and Elevation Array Switching Unit 8, as are the signals from antenna Arrays 3 and 4, after which the two sums are separately inputted to Receivers 13 and 14, respectively, representing inputs to two separate synthetic arrays for elevation interferometric phase comparison.”; Fig. 7, Range-azimuth map generator 26 is downstream of azimuth and elevation array switching unit 8).
Di, Aydogdu, Baker, and Boles are considered to be analogous to the claimed invention because they are in the same field of radar-based occupancy determination. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Di in view of Aydogdu and further in view of Baker with the teachings of Boles with the motivation of increasing data SNR, filtering noise, and thus improving weak target detection.
Conclusion
The prior art made of record and not relied upon is considered pertinent to Applicant's disclosure.
Huang et al. (RPCRS: Human Activity Recognition Using Millimeter Wave Radar [2023]) teaches a radar system that removes detected points of a point cloud outside of a region of interest, the remaining points processed to classify human activity, the point cloud indicating range and angle (see section III pages 124-125).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ERIC K HODAC whose telephone number is (571) 270-0123. The examiner can normally be reached M-Th 8-6.
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.
/ERIC K HODAC/Examiner, Art Unit 3648
/OLUMIDE AJIBADE AKONAI/Primary Examiner, Art Unit 3648
Prosecution Insights