DETAILED ACTION
Response to Amendments
Amendment filed on December 19, 2025 has been entered.
Claims 1, 8, and 15 are amended.
Claim 3, 10, and 19 is cancelled.
Claims 1-2, 4-9, 11-18 and 20 are pending this application.
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 1-16 are rejected under 35 U.S.C. 103 as being unpatentable over Anushiravani et al (US 11,055,575 B2) in view of Gyer et al (EP 3 835 812 A1) and further in view of Tsai et al (arXiv: 2019).
Regarding Claim 1, Anushiravani teaches an electronic apparatus comprising [col 26, lines 35-45]:
a sensor [col 26, lines 35-45];
a memory storing at least one instruction [col 26, lines 35-45];
and a processor connected to the memory, the instructions, when executed by the processor cause the electronic apparatus to [col 26, lines 35-45]:
transmit a radar signal through the sensor and receive a signal reflected by a user [col 26, lines 45-67 for using signals and capturing data],
based on the signal reflected by the user [col 26, lines 45-67 capturing data],
acquire first information on a user's movement [col 26, lines 45-67, col 27, lines 15-25 for using motion sensors (user’s movement)],
acquire second information on the user's movement by performing a Fourier transform on the first information [col 28, lines 10-20],
acquire first feature information corresponding to the first information by inputting the first information into a first neural network [col 26, lines 45-65 for putting symptoms (first feature) into a first neural network],
acquire second feature information corresponding to the second information by inputting the second information into the second neural network [col 26, lines 45-67 for second feature disease classifier into second neural network],
acquire sleep information on a user's sleep by inputting the input data into a third neural network [col 26, lines 45-67 and col 27, lines 1-30],
and provide the sleep information on the user’s sleep [col 26, lines 45-67]
wherein the first information includes a distance map with respect to a movement of a user's chest [col 24, lines 55-67 for sensing cough and wheezing].
Anushiravani fails to explicitly teach an ultra-wideband (UWB) sensor and the second information includes a Doppler map with respect to the movement of the user's chest.
Gyger has localization of living individuals in a search area (abstract) and teaches an ultra-wideband (UWB) sensor [0021-0022]
and the second information includes a Doppler map with respect to the movement of the user's chest [0027-0030, 0051-0052]
and wherein the processor is further configured to acquire input data acquired by performing position embedding on feature information and combining element-wise the first feature information and the second feature information [0051-0052].
It would have been obvious to a person of ordinary skill in the art before the effective filling date of the applicant’s invention for modifying the sleep monitoring techniques, as disclosed by Anushiravani, further including the sensor calculations as taught by Gyger for the purpose to identify several types of vital signs (Gyger, 0028).
Anushiravani fails to explicitly teach acquire input data acquired by performing position embedding on feature information combining element-wise the first feature information and the second feature information.
Tsai has a method to generically address the above issues in an end-to-end manner without explicitly aligning the data [page 1, abstract] and teaches acquire input data acquired by performing position embedding on feature information combining element-wise the first feature information and the second feature information [page 3, figure 2 for using positional embedding of different transformer (features) with page 4, left column last paragraph].
It would have been obvious to a person of ordinary skill in the art before the effective filling date of the applicant’s invention for modifying the sleep monitoring techniques, as disclosed by Anushiravani, further including the embedding calculations as taught by Tsai for the purpose to result in low-level position aware features for different modalities (Tsai, page 4, right column, first paragraph).
Regarding Claim 8, Anushiravani teaches a method of controlling an electronic apparatus, the method comprising [col 26, lines 35-45]:
transmitting a radar signal through a sensor and receiving a signal reflected by a user [col 26, lines 35-45];
based on the signal reflected by the user [col 26, lines 40-67],
acquiring first information on a user's movement [col 26, lines 45-67, col 27, lines 15-25 for using motion sensors (user’s movement)];
acquiring second information on the user's movement by performing a Fourier transform on the first information [col 26, lines 45-67, col 28, lines 1-20];
acquiring first feature information corresponding to the first information by inputting the first information into a first neural network [col 26, lines 45-67, col 28, lines 1-20];
acquiring second feature information corresponding to the second information by inputting the second information into a second neural network [col 26, lines 40-67];
acquiring sleep information on a user's sleep by inputting input data into a third neural network [col 26, lines 40-67];
and providing the sleep information on the user’s sleep [col 26, lines 45-67]
wherein the first information includes a distance map with respect to a movement of a user's chest [col 24, lines 55-67 for sensing cough and wheezing].
Anushiravani fails to explicitly teach an ultra-wideband (UWB) sensor and the second information includes a Doppler map with respect to the movement of the user's chest.
Gyger has localization of living individuals in a search area (abstract) and teaches an ultra-wideband (UWB) sensor [0021-0022]
and the second information includes a Doppler map with respect to the movement of the user's chest [0027-0030]
and wherein the processor is further configured to acquire input data acquired by performing position embedding on feature information and combining element-wise the first feature information and the second feature information [0051-0052].
It would have been obvious to a person of ordinary skill in the art before the effective filling date of the applicant’s invention for modifying the sleep monitoring techniques, as disclosed by Anushiravani, further including the sensor calculations as taught by Gyger for the purpose to identify several types of vital signs (Gyger, 0028).
Anushiravani fails to explicitly teach acquire input data acquired by performing position embedding on feature information combining element-wise the first feature information and the second feature information.
Tsai has a method to generically address the above issues in an end-to-end manner without explicitly aligning the data [page 1, abstract] and teaches acquire input data acquired by performing position embedding on feature information combining element-wise the first feature information and the second feature information [page 3, figure 2 for using positional embedding of different transformer (features) with page 4, left column last paragraph].
It would have been obvious to a person of ordinary skill in the art before the effective filling date of the applicant’s invention for modifying the sleep monitoring techniques, as disclosed by Anushiravani, further including the embedding calculations as taught by Tsai for the purpose to result in low-level position aware features for different modalities (Tsai, page 4, right column, first paragraph).
Regarding Claim 15, Anushiravani teaches one or more non-transitory computer-readable storage media storing one or more computer programs including that, when executed by a processor of an electronic apparatus, cause the electronic apparatus to perform operations, the operations comprising [col 26, lines 35-45]:
transmitting a radar signal through a sensor and receiving a signal reflected by a user [col 26, lines 35-45];
based on the signal reflected by the user [col 26, lines 35-55],
acquiring first information on a user's movement [col 26, lines 45-67, col 27, lines 15-25 for using motion sensors (user’s movement)];
acquiring second information on the user's movement by performing a Fourier transform on the first information [col 26, lines 45-67, col 28, lines 1-20];
acquiring first feature information corresponding to the first information by inputting the first feature information into a first neural network [col 26, lines 45-67, col 28, lines 1-20];
acquiring second feature information corresponding to the second information by inputting the second information into a second neural network [col 26, lines 45-67];
acquiring sleep information on a user's sleep by inputting input data into a third neural network [col 26, lines 45-67];
and providing the sleep information on the user’s sleep [col 26, lines 45-67]
wherein the first information includes data with respect to a movement of a user's chest [col 24, lines 55-67 for sensing cough and wheezing].
Anushiravani fails to explicitly teach an ultra-wideband (UWB) sensor and the second information includes a Doppler map with respect to the movement of the user's chest.
Gyger has localization of living individuals in a search area (abstract) and teaches an ultra-wideband (UWB) sensor [0021-0022]
and the second information includes a Doppler map with respect to the movement of the user's chest [0027-0030]
and wherein the processor is further configured to acquire input data acquired by performing position embedding on feature information and combining element-wise the first feature information and the second feature information [0051-0052].
It would have been obvious to a person of ordinary skill in the art before the effective filling date of the applicant’s invention for modifying the sleep monitoring techniques, as disclosed by Anushiravani, further including the sensor calculations as taught by Gyger for the purpose to identify several types of vital signs (Gyger, 0028).
Anushiravani fails to explicitly teach acquire input data acquired by performing position embedding on feature information combining element-wise the first feature information and the second feature information.
Tsai has a method to generically address the above issues in an end-to-end manner without explicitly aligning the data [page 1, abstract] and teaches acquire input data acquired by performing position embedding on feature information combining element-wise the first feature information and the second feature information [page 3, figure 2 for using positional embedding of different transformer (features) with page 4, left column last paragraph].
It would have been obvious to a person of ordinary skill in the art before the effective filling date of the applicant’s invention for modifying the sleep monitoring techniques, as disclosed by Anushiravani, further including the embedding calculations as taught by Tsai for the purpose to result in low-level position aware features for different modalities (Tsai, page 4, right column, first paragraph).
Regarding Claim 2 and 9, Anushiravani teaches the instructions, when executed by the processor, further cause the electronic apparatus to [col 26, lines 35-67]:
acquire the sleep information on the user's sleep by inputting a value acquired by multiplying the first feature information and the second feature information for each element [col 26, lines 35-67]
into the third neural network [col 26, lines 35-67].
Regarding Claim 3 and 10, Anushiravani fails to explicitly teach the first information includes a distance map with respect to a movement of a user’s chest, and wherein the second information includes a Doppler map with respect to the movement of the user's chest.
Gyger has localization of living individuals in a search area (abstract) and teaches the first information includes a distance map with respect to a movement of a user’s chest [0052-0056],
and wherein the second information includes a Doppler map with respect to the movement of the user's chest [0027-0030, 0052-0056].
It would have been obvious to a person of ordinary skill in the art before the effective filling date of the applicant’s invention for modifying the sleep monitoring techniques, as disclosed by Anushiravani, further including the sensor calculations as taught by Gyger for the purpose to identify several types of vital signs (Gyger, 0028).
Regarding Claim 4 and 11, Anushiravani teaches the first information includes time series information on the user's movement according to a change of time [Figure 26, element 2601, col 20, lines 15-40],
and wherein the second information includes time series information on a frequency of the user's movement according to the change of time [figure 26].
Regarding Claim 5 and 12, Anushiravani teaches the first neural network includes a convolutional neural network (CNN) for outputting the first feature information [col 14, lines 30-55],
and wherein, the instructions, when executed by the processor, further cause the electronic apparatus to [col 14, lines 30-55]:
acquire restored first information using the first feature information, and train parameters of the CNN by comparing the restored first information with the first information [col 14, lines 35-60].
Regarding Claim 6 and 13, Anushiravani teaches the second neural network includes a convolutional neural network (CNN) for outputting the second feature information, and wherein, the instructions, when executed by the processor, further cause the electronic apparatus to [col 7, line 52 to col 8, line 30]:
acquire restored second information by using the second feature information, and train parameters of the CNN by comparing the restored second information with the second information [col 8, lines 1-30 and col 8, lines 40-55].
Regarding Claim 7 and 14, Anushiravani teaches the instructions, when executed by the processor, further cause the electronic apparatus to: train the first neural network, the second neural network, and the third neural network based on movement information on the user's movement measured over a predetermined time [col 26, lines 30-67].
Regarding Claim 16, Anushiravani teaches the acquiring of the sleep information includes inputting data [col 18, lines 1-20],
which is acquired by performing position embedding on feature information combining the first feature information and the second feature information, into the third neural network [col 18, lines 10-30 and col 26, lines 50-67].
Claim 17 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Anushiravani et al (US 11,055,575 B2) in view of Gyer et al (EP 3 835 812 A1) and Tsai et al (arXiv: 2019), as applied to Claims 8 and 15 above, and further in view An et al (ArXiv, 2018).
Regarding Claim 17, Anushiravani teaches the first neural network includes a first Conv1d layer, a first batch normalization layer, [figure 3B element A1 for first convolved layer, col9, lines 40-60],
wherein the second neural network includes a second Conv1d layer, a second batch normalization layer, [figure 3B element A1 for first convolved layer, col9, lines 40-60],
and wherein the third neural network includes at least one of a Bidirectional-Long Short-Term Memory (Bi-LSTM) layer or a Transformer network layer [col 16, lines 50-67].
Anushiravani fails to explicitly teaches a first Squeeze and Excitation network (SENet) layer and a second SENet layer.
An has Squeeze and-Excitation Long-term Recurrent Convolutional Networks
(SE-LRCN) for human action recognition (page 1, left column, abstract) and teaches a first Squeeze and Excitation network (SENet) layer and a second SENet layer [page 1, right column, first paragraph].
It would have been obvious to a person of ordinary skill in the art before the effective filling date of the applicant’s invention for modifying the sleep monitoring techniques, as disclosed by Anushiravani, further including the human recognition calculations as taught by An for the purpose to implement feature recalibration (An, page 1, right column, first paragraph).
Regarding Claim 20, Anushiravani fails to explicitly teach the first neural network, the second neural network, and the third neural network have a structure of a two stream one dimensional convolution neural network (lDCNN)-squeeze and excitation network (SENet) - transformer network.
An has Squeeze and-Excitation Long-term Recurrent Convolutional Networks
(SE-LRCN) for human action recognition (page 1, left column, abstract) and teaches the first neural network, the second neural network, and the third neural network have a structure of a two stream one dimensional convolution neural network (lDCNN)-squeeze and excitation network (SENet) - transformer network [page 1, right column, first paragraph and page 5, Table V].
It would have been obvious to a person of ordinary skill in the art before the effective filling date of the applicant’s invention for modifying the sleep monitoring techniques, as disclosed by Anushiravani, further including the human recognition calculations as taught by An for the purpose to implement feature recalibration (An, page 1, right column, first paragraph).
Claim 18 is rejected under 35 U.S.C. 103 as being unpatentable over Anushiravani et al (US 11,055,575 B2) in view of Gyer et al (EP 3 835 812 A1) and Tsai et al (arXiv: 2019), as applied to Claim 15 above, and further in view Nakata et al (US 20140/194793 A1).
Regarding Claim 18, Anushiravani fails to explicitly teach in providing the sleep information on the user’s sleep, the control method further comprises: in response to acquiring the sleep information, displaying a user's sleep apnea or hypopnea diagnosis result, which includes information on one of a mild degree, a moderate degree, or a severe degree of the user's sleep apnea or hypopnea, together with a user's Apnea-Hypopnea Index (AHI) index information.
Nakata has a non-contact physiological motion sensor and a monitor device that can incorporate use of the Doppler effect (abstract) and teaches in providing the sleep information on the user’s sleep, the control method further comprises [0024]:
in response to acquiring the sleep information, displaying a user's sleep apnea or hypopnea diagnosis result [0023-0025 smartphone, tablet],
which includes information on one of a mild degree, a moderate degree, or a severe degree of the user's sleep apnea or hypopnea [0023-0025 for sleep quality and determining severity of apnea],
together with a user's Apnea-Hypopnea Index (AHI) index information [0024].
It would have been obvious to a person of ordinary skill in the art before the effective filling date of the applicant’s invention for modifying the sleep monitoring techniques, as disclosed by Anushiravani, further including the sensor calculations as taught by Nakata for the purpose to assist the clinician in assessing subject's apnea severity by reporting sleep breathing disorder events (Nakata, 0024).
Response to Arguments
Applicant’s arguments with respect to claims 1-2, 4-9, 11-18 and 20 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
In applicant’s remarks, page 8 first paragraph, applicant argues that the reference does not disclose acquiring second information on user’s move by using a Fourier transform on the first information. Examiner respectfully disagrees: Gyger also teaches this feature by using a 2D FFT to create rand doppler maps and using radar to detect breathing and chest movements [Gyger, 0027-0030, 0052].
In applicant’s remarks, page 8 last paragraph, applicant argues that the audio signal is used to train the neural network, not input in the neural network. Examiner respectfully disagrees: Training a neural network requires inputting signals for training [Anushiravani, col 6, line 57 to col 7, lines 20].
In applicant’s remarks, page 9, third paragraph, applicant argues that the reference does not disclose the first neural network and second neural network in the structure recited in the independent claims. Examiner respectfully disagrees: Anushiravani teaches using multiple neural networks including CNNs to process data streams and teaching using a Fourier transform for frequency domain analysis [Anushiravani, claim 1, and figure 11-12B].
In applicant’s remarks, page 9, last paragraph, applicant argues that the reference does not disclose the user’s movement in claim 8. Examiner respectfully disagrees: The first neural network extracts features from the time domain sensor such as respiratory sounds, the second neural network extracts frequency domain feature using a Fourier transform with the signal, the sensor type is merely a design choice [Anushiravani, claim 1, and claim 5].
In applicant’s remarks, page 11, second paragraph, applicant argues that Gyger does not remedy data with the movement of the user’s chest and a Doppler map. Examiner respectfully disagrees: Gyger teaches breathing and heartbeats as well as using range Doppler maps to detect vital signs [Gyger, 0025, and 0027-0030].
Conclusion
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 nonprovisional extension fee (37 CFR 1.17(a)) 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 SAMARINA MAKHDOOM whose telephone number is (703)756-1044. The examiner can normally be reached Monday – Thursdays from 8:30 to 5:30 pm eastern time.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, William Kelleher can be reached on 571-272-7753 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.
/SAMARINA MAKHDOOM/
Examiner, Art Unit 3648