# US 12,183,119 B2 · Face identification using millimeter-wave radar sensor data > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,183,119 B2 - Document ID: US12183119B2 - Status: Granted - Kind: B2 - Application number: 17/575,512 - Filed: 2022-01-13 - Published: 2022-10-13 - Granted: 2024-12-31 - Earliest claimed: 2021-04-12 - Assignees: Bitsensing Inc. - Applicants: Bitsensing Inc. - Inventors: Jae-Eun Lee; Haesueng Lim; Seongwook Lee - Web page: https://patentgrep.com/patent/US12183119B2 ## CPC - G06V 40/172: {Classification, e.g. identification} - G01S 7/417: {involving the use of neural networks} - G01S 13/34: using transmission of continuous, frequency-modulated waves while heterodyning the received signal, or a signal derived therefrom, with a locally-generated signal related to the contemporaneously transmitted signal - G01S 13/89: for mapping or imaging - G06V 10/82: using neural networks - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING - G01S: RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES ## Abstract Facial identification of subjects using a millimeter wave radar sensor, including: transmitting a radar signal onto faces of subjects using the millimeter wave radar sensor; receiving and accumulating reflected radar data, wherein the reflected radar data is received from channels of multiple receiving antennas; converting the reflected radar data into an image form suitable as an input to a convolutional neural network; and performing the facial identification on the image form using a classifier trained by the convolutional neural network. ## Claims ### Claim 1 A method for facial identification of subjects using a millimeter wave radar sensor, the method comprising: transmitting a radar signal onto faces of subjects using the millimeter wave radar sensor; receiving and accumulating reflected radar data, wherein the reflected radar data is received from channels of multiple receiving antennas; converting the reflected radar data into an image form as an input to a convolutional neural network; and performing the facial identification on the image form using a classifier trained by the convolutional neural network, wherein the convolutional neural network comprises: a first number of convolution layers; and wherein the first number is determined by a number of the faces of the subjects to identify. ### Claim 2 (depends on 1) The method of claim 1, wherein the millimeter wave radar sensor includes a frequency-modulated continuous-wave radar sensor. ### Claim 3 (depends on 2) The method of claim 2, wherein the frequency-modulated continuous-wave radar sensor is configured with an operating center frequency of 61 GHz and a bandwidth of 6 GHz. ### Claim 4 (depends on 1) The method of claim 1, wherein converting the reflected radar data into an image form comprises synthesizing in parallel the reflected radar data received from the channels of the multiple receiving antennas using fast Fourier Transform. ### Claim 5 (depends on 1) The method of claim 1, wherein the classifier uses only the radar data around the faces of the subjects. ### Claim 6 (depends on 1) The method of claim 1, wherein the convolutional neural network further comprises: a second number of fully-connected layers, wherein the second number is determined by a number of the faces of the subjects to identify. ### Claim 7 (depends on 6) The method of claim 6, wherein each of the first number of convolutional layers performs n×n convolutions, wherein n=the first number. ### Claim 8 (depends on 7) The method of claim 7, further comprising flattening features of a last layer of the first number of convolutional layers by at least one fully-connected layer. ### Claim 9 A system to identify faces of subjects, comprising: at least one transmitting antenna to transmit a radar signal onto faces of subjects; a plurality of receiving antenna elements to receive and accumulate reflected radar data; and a digital signal processor to convert the reflected radar data into an image form as an input to a convolutional neural network, the digital signal processor to perform the facial identification on the image form using a classifier trained by the convolutional neural network, wherein the convolutional neural network comprises: a first number of convolutional layers, and wherein the first number is determined by a number of the faces of the subjects to identify. ### Claim 10 (depends on 9) The system of claim 9, wherein the system is a frequency-modulated continuous wave (FMCW) radar sensor. ### Claim 11 (depends on 10) The system of claim 10, wherein the FMCW radar sensor is configured as a millimeter wave radar sensor with a center frequency of 61 GHz and a bandwidth of 6 GHz. ### Claim 12 (depends on 9) The system of claim 9, wherein the system is configured with one transmitting antenna and three receiving antenna elements, wherein two receiving antenna elements are placed in a horizontal direction, while two receiving antenna elements are placed in a vertical, direction, with one receiving antenna element is shared in the horizontal and vertical directions. ### Claim 13 (depends on 9) The system of claim 9, further comprising a waveform generator to transmit a waveform whose frequency increases linearly. ### Claim 14 (depends on 9) The system of claim 9, wherein the convolutional neural network further comprises: a second number of fully-connected layers, wherein the second number is determined by a number of the faces of the subjects to identify. ### Claim 15 A non-transitory computer-readable storage medium storing a computer program to identify faces of subjects, the computer program comprising executable instructions that cause a computer to: command transmission of a radar signal onto faces of subjects using a millimeter wave radar sensor; receive and accumulate reflected radar data, wherein the reflected radar data is received from channels of multiple receiving antennas; convert the reflected radar data into an image form as an input to a convolutional neural network; and perform facial identification on the image form using a classifier trained by the convolutional neural network, wherein the convolutional neural network comprises: a first number of convolutional layer, and wherein the first number is determined by a number of the faces of the subjects to identify. ### Claim 16 (depends on 15) The computer-readable storage medium of claim 15, wherein the executable instructions that cause the computer to convert the reflected radar data into an image form comprises executable instruction that cause the computer to synthesize in parallel the reflected radar data received from the channels of the multiple receiving antennas using fast Fourier Transform. ### Claim 17 (depends on 15) The computer-readable storage medium of claim 15, wherein the classifier uses only the radar data around the faces of the subjects. ### Claim 18 (depends on 15) The computer-readable storage medium of claim 15, wherein the convolutional neural network further comprises: a second number of fully-connected layers, wherein the second number is determined by a number of the faces of the subjects to identify. ### Claim 19 (depends on 18) The computer-readable storage medium of claim 18, wherein each of the first number of convolutional layers performs n×n convolutions, wherein n=the first number. ### Claim 20 (depends on 19) The computer-readable storage medium of claim 19, further comprising executable instructions that cause the computer to flatten features of a last layer of the first number of convolutional layers by at least one fully-connected layer.