Method for operating a sensor device, sensor device, and electronic device

Abstract

A method for operating a sensor device comprises recording an array of first values as a first dataset and an array of second values as a second dataset. Recording a first value includes emitting light having a first wavelength and detecting an intensity of reflected light having the first wavelength as the first value. Recording a second value includes emitting light having a second wavelength and detecting an intensity of reflected light having the second wavelength as the second value. The method further includes classifying a state of the sensor device in dependence of the first dataset and the second dataset.

Description

[0001] 2024PF01558 1

[0002] METHOD FOR OPERATING A SENSOR DEVICE , SENSOR DEVICE , AND ELECTRONIC DEVICE

[0003] TECHNICAL FIELD

[0004] The present invention relates to a method for operating a sensor device, to a sensor device, and to an electronic module .

[0005] This patent application claims the priority of US provisional patent application US 63 / 737, 829, the disclosure content of which is hereby incorporated by reference .

[0006] BACKGROUND

[0007] Proximity devices for detecting an approach of the proximity device to an obj ect are known in the state of the art .

[0008] It is an obj ect of the present invention to provide a method for operating a sensor device . It is a further obj ect of the present invention to provide a sensor device . It is a further obj ect of the present invention to provide an electronic device . These obj ectives are achieved by a method for operating a sensor device, by a sensor device, and by an electronic device according to the independent claims . Different variants are disclosed in the dependent claims .

[0009] According to one aspect, a method for operating a sensor device comprises recording an array of first values as a first dataset and an array of second values as a second dataset . Recording a first value includes emitting light having a first wavelength and detecting an intensity of reflected light having the first wavelength as the first value . Recording a second value includes emitting light having a second wavelength and detecting an intensity of reflected light having the second wavelength as the second value . The method further includes classifying a state of the sensor device in dependence of the first dataset and the second dataset . This2024PF01558 2

[0010] method allows a fast and reliable determination of the state of the sensor device using only optical measurements .

[0011] In some variants, classifying the state comprises a classification of whether the sensor device is facing skin. This may allow a determination of whether a portable or wearable device is currently employed by a user of the device .

[0012] In some variants, recording the array of first values and the array of second values is started when the first values exceed a first intensity threshold value or the second values exceed a second intensity threshold value . Exceeding the first intensity threshold value or the second intensity threshold value may indicate that the state of the sensor device has changed or is currently changing.

[0013] In some variants, the array of first values and the array of second values is recorded over a pre-defined period of time between 0.5 seconds and 5 seconds . Datasets of such length have proven to be useful for classifying the state of the sensor device .

[0014] In some variants, the array of first values and the array of second values is recorded over a pre-defined period of time between 1 second and 3 seconds . Datasets of this length have proven to be useful for classifying the state of the sensor device .

[0015] In some variants, the state of the sensor device is classified in dependence of a temporal development of the first values and the second values . This may take advantage of a change of a spatial location of the sensor device being reflected as a temporal development in the first values and the second values .

[0016] In some variants, classifying the state of the sensor device includes filtering the array of first values and the array of second values using a zero-phase linear filter . Filtering the2024PF01558 3

[0017] array of first values and the array of second values may reduce noise in the first dataset and the second dataset and may allow for a more robust classification of the state of the sensor device in dependence of the first dataset and the second dataset .

[0018] In some variants, classifying the state of the sensor device includes calculating a first derivative of the first dataset and a first derivative of the second dataset, calculating a variance of the first derivative of the first dataset as a first variance and calculating a variance of the first derivative of the second dataset as a second variance, calculating an absolute difference between the first variance and the second variance, calculating an absolute sum of the first variance and the second variance, and calculating a feature as the ratio between the absolute difference and the absolute sum. The state of the sensor device is classified in dependence of this feature . This may allow a simple and reliable classification of the state of the sensor device .

[0019] In some variants, classifying the state of the sensor device includes calculating other or further features . Examples of possible features and definitions for calculating them are disclosed in the detailed description below. In some variants, the state of the sensor device is classified in dependence of one or more of these other features .

[0020] In some variants, two or more different features are combined to form further features . In some variants, the state of the sensor device is classified in dependence of one or more of these further features .

[0021] In some variants, classifying the state of the sensor device may include comparing one or more features with one or more threshold values .

[0022] In some variants, the state of the sensor device is classified as not facing skin if the feature is smaller than or2024PF01558 4

[0023] equal to a first state threshold value . The state of the sensor device is classified as facing skin if the feature is larger than or equal to a second state threshold value . The state of the sensor device is classified as indeterminate if the feature is between the first state threshold value and the second state threshold value . This allows a fast and simple classification of the state of the sensor device .

[0024] In some variants, the state of the sensor device is classified in dependence of one or more features using a trained machine learning model . This may allow a reliable and robust classification of the state of the sensor device .

[0025] In some variants, the state of the sensor device is classified in dependence of one or more features using a trained neural network.

[0026] According to a further aspect, a method for operating a sensor device comprises recording an array of first values as a first dataset, wherein recording a first value includes emitting light having a first wavelength, and detecting an intensity of reflected light having the first wavelength as the first value . The method further includes classifying a state of the sensor device in dependence of the first dataset . This method also allows a fast and reliable determination of the state of the sensor device using only optical measurements .

[0027] In some variants, recording the array of first values is started when the first values exceed a first intensity threshold value . Exceeding the first intensity threshold value may indicate that the state of the sensor device has changed or is currently changing.

[0028] In some variants, classifying the state of the sensor device includes filtering the array of first values using a zerophase linear filter . Filtering the array of first values may reduce noise in the first dataset and may allow for a more2024PF01558 5

[0029] robust classification of the state of the sensor device independent of the first dataset .

[0030] According to a further aspect, a sensor device comprises an emission unit for emitting light having a first wavelength and emitting light having a second wavelength, and a detection unit for detecting an intensity of reflected light having the first wavelength and detecting an intensity of reflected light having the second wavelength. The sensor device is configured to carry out a method as disclosed above . This sensor device allows a fast and reliable determination of a state of the sensor device using optical measurements carried out using the emission unit and the detection unit of the sensor device .

[0031] According to a further aspect, an electronic device comprises a sensor device as disclosed above . The sensor device of the electronic device may allow a determination of a state of the electronic device .

[0032] In some variants, the electronic device may be an earbud. The state of the electronic device may include a location of the earbud, for example a location of the earbud in the ear of a human user .

[0033] BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The above-described properties, features, and advantages of the invention, as well as the manner in which they are achieved, will become more clearly and comprehensively understandable through the following description of exemplary variants . These variants will be explained in more detail in conjunction with the drawings, in which, in schematic representation :

[0035] Fig. 1 shows a top view of an electronic device comprising a sensor device;2024PF01558 6

[0036] Fig. 2 shows a schematic side view of the electronic device during an approach to a target;

[0037] Fig. 3 shows a schematic temporal development of a first intensity and a second intensity during an approach of the sensor device to a target;

[0038] Fig. 4 shows a further schematic temporal development of the first intensity and the second intensity during an approach of the electronic device to a target;

[0039] Fig. 5 shows a further schematic depiction of a temporal development of the first intensity and the second intensity during an approach of the electronic device to a target;

[0040] Fig. 6 shows various features extracted from temporal developments of the first intensity and the second intensity; and

[0041] Fig. 7 shows an exemplary depiction of two features in two different states of the sensor device .

[0042] DETAILED DESCRIPTION

[0043] Fig. 1 shows a schematic top view of an electronic device 10. Fig. 2 shows a schematic sectional view of the electronic device 10. In one variant, the electronic device 10 is an earbud (earphone) 11. In other variants, the electronic device 10 may be a mobile phone, a wearable device such as a watch or a fingerring, or another type of electronic device .

[0044] The electronic device 10 comprises a sensor device 20 that is integrated into the electronic device 10. The sensor device 20 has a detection side 21 which may form a part of an outer surface of the electronic device 10.

[0045] The sensor device 20 is designed for classifying a state of the sensor device 20. The sensor device 20 may be designed to detect an approximation of the detection side 21 of the2024PF01558 7

[0046] sensor device 20 to a target 30 or to detect an arrangement of the electronic device 10 in which the detection side 21 of the sensor device 20 is oriented towards the target 30 and is in proximity to the target 30. The sensor device 20 may also be designed to detect if the target 30 is skin of a human body .

[0047] The sensor device 20 comprises an emission unit 100 and a detection unit 200.

[0048] The emission unit 100 is designed for emitting first light 111 having a first wavelength and emitting second light 121 having a second wavelength that is different from the first wavelength. In one variant, the first wavelength is 850 nm and the second wavelength is 940 nm. In another variant, the first wavelength is 940 nm and the second wavelength is 850 nm. In further variants, one or both of the first wavelength and the second wavelength are different from 850 nm and 940 nm.

[0049] In the example depicted in Figs . 1 and 2, the emission unit comprises a first emitter 110 for emitting the first light 111 and a second emitter 120 for emitting the second light 121. This may allow the first emitter 110 and the second emitter 120 to operate independently from each other such that either only the first light 111 or only the second light 121 is emitted or both the first light 111 and the second light 121 are emitted simultaneously. In other variants, however, the emission unit 100 may comprise a common emitter for emitting both the first light 111 and the second light 121 simultaneously .

[0050] The detection unit 200 is designed for detecting a first intensity 211 of light having the first wavelength and a second intensity 221 of light having the second wavelength. In the example depicted in Figs . 1 and 2, the detection unit 200 comprises a first detector 210 for detecting the first intensity 211 of light having the first wavelength and a second2024PF01558 8

[0051] detector 220 for detecting the second intensity 221 of light having the second wavelength. In other variants, however, the detection unit 200 may comprise a common detector for detecting both the first intensity 211 of light having the first wavelength and the second intensity 221 of light having the second wavelength.

[0052] In operation of the sensor device 20, the emission unit 100 emits the first light 111 having the first wavelength and the second light 121 having the second wavelength simultaneously or alternately one after the other . The emission unit 100 and the detection 200 are arranged such that no or only little of the first light 111 and the second light 120 emitted from the emission unit 100 reaches the detection unit 200 directly. Only first light 111 and second light 121 that is reflected at the target 30 reaches the detection unit 200 and is detected as the first intensity 211 and the second intensity 221, respectively.

[0053] Fig. 3 shows a schematic depiction of a temporal development of the first intensity 211 and the second intensity 221 measured over a period of time 300 during an approach of the detection side 21 of the sensor device 20 towards the target 30. In this example, the electronic device 10 is an earbud 11 and is inserted into the ear of a user such that the target 30 is composed of skin.

[0054] At the start of the period of time 300 depicted in Fig. 3, both the first intensity 211 and the second intensity 221 are low because the sensor device 20 is far away from the target 30 and none of the first light 111 and the second light 121 is reflected towards the detection unit 200. The first intensity 211 and the second intensity 221 sharply rise during an approach phase while the user presses the earbud into the desired position in the ear . The sharp rise of the first intensity 211 and the second intensity 221 is followed by a fuss phase in which the user moves the earbud 11 into the best position and the first intensity 211 and the second intensity2024PF01558 9

[0055] 221 experience several oscillations . This is followed by a rebound phase after the user lets go of the earbud 11 and it retracts slightly. In the rebound phase, the first intensity 211 and the second intensity 221 may decrease before reaching a relatively static level . Towards the end of the depicted period of time 300, the earbud 11 is removed from the ear of the user again and the first intensity 211 and the second intensity 221 sharply drop .

[0056] Each of the approach phase, the fuss phase and the rebound phase may comprise characteristic patterns that allow to distinguish the case in which the earbud 11 is inserted into an ear from other cases in which the earbud 11 is inserted into a case, put in a pocket, put on a table, held in a hand, or handled in some other way.

[0057] Furthermore, the levels and temporal developments of the first intensity 211 and the second intensity 221 may also allow to distinguish cases in which the target 30 is skin from other cases in which the target 30 is comprised of a different material . Fig. 4 shows an exemplary depiction of a temporal development of the first intensity 211 and the second intensity 221 during an approach of the detection side 21 of the sensor device 20 of the electronic device 10 towards the target 30 in a case in which the target 30 is a grey card. Fig. 5 shows a schematic depiction of a temporal development of the first intensity 211 and the second intensity 221 during an approach of the detection side 21 of the sensor device 20 towards the target 30 in a case in which the target 30 is skin. A comparison of Figs . 4 and 5 shows that the temporal development of the first intensity 211 and the second intensity 221 allows to distinguish the cases in which the target 30 is skin from the cases in which the target 30 is not skin. In the depicted example, the first intensity 211 and the second intensity 221 drop towards the end of the depicted period of time during the approach phase in the case in which the target 30 is a grey card but remain comparatively high in the case in which the target 30 is skin.2024PF01558 10

[0058] The sensor device 20 is designed to record an array of first values PD1 as a first dataset PDl_Xsec and to record an array of second values PD2 as a second dataset PD2_Xsec . Recording a first value PD1 includes emitting the first light 111 having the first wavelength using the emission unit 100 and detecting the first intensity 211 of reflected light having the first wavelength as the first value PD1 using the detection unit 200. Recording a second value PD2 in each case includes emitting the second light 121 having the second wavelength using the emission unit 100 and detecting the second intensity 221 of reflected light having the second wavelength as the second value PD2 using the detection unit 200.

[0059] In some variants, recording the array of first values PD1 as the first dataset PDl_Xsec and the array of second values PD2 as the second dataset PD2_Xsec is started when the first values PD1 (the first intensity 211 ) exceed a pre-defined first intensity threshold value or when the second values PD2 (the second intensity 221 ) exceed a pre-defined second intensity threshold value or when both the first intensity 211 and the second intensity 221 exceed the respective intensity threshold values .

[0060] In some variants, the first dataset PDl_Xsec (the array of first values PD1 ) and the second dataset PD2_Xsec (the array of second values PD2 ) is recorded over a pre-defined period of time between 0.5 seconds and 5 seconds . In some variants, the pre-defined period of time is between 1 second and 3 seconds . In some variants, the pre-defined period of time is 2 seconds .

[0061] It is convenient that both the first dataset PDl_Xsec and the second dataset PD2_Xsec comprise the same length N. This means that the first dataset PDl_Xsec comprises N first values PD1 and the second dataset PD2_Xsec comprises N second values PD2 . The length N may be 256, 1024, 2000 or any other number .2024PF01558 11

[0062] Elements of the first dataset PDl_Xsec and the second dataset PD2_Xsec having the same index will be referred to as corresponding elements in the following. For example, the second element of the first dataset PDl_Xsec corresponds to the second element of the second dataset PD2_Xsec . Corresponding elements of the first dataset PDl_Xsec and the second dataset PD2_Xsec are recorded simultaneously or immediately after each other .

[0063] The sensor device 20 is designed for classifying a state of the sensor device 20 in dependence of the first dataset PDl_Xsec and the second dataset PD2_Xsec . In particular, the state of the sensor device 20 may be classified in dependence of a temporal development of the first values PD1 of the first dataset PDl_Xsec and the second values PD2 of the second dataset PD2_Xsec . Classifying the state of the sensor device 20 may include a classification of whether the sensor device 20 is facing skin. The sensor device 20 may be capable to detect if the detection side 21 of the sensor device 20 faces a target 30 and whether the target 30 is skin.

[0064] The sensor device 20 may be designed to filter one or both of the recorded first dataset PDl_Xsec and second dataset PD2_xsec using a filter such as a zero-phase linear filter or another kind of filter . In this case, the filtered datasets are used instead of the first dataset PDl_Xsec and the second dataset PD2_Xsec, respectively, for classifying the state of the sensor device 20. This also applies to the calculation of the features described in the following.

[0065] For the classification of the state of the sensor device 20, it is possible to extract one or more features from the first dataset PDl_Xsec and the second dataset PD2_Xsec that are characteristic for the state of the sensor device 20.

[0066] Fig. 6 shows exemplary scatter plots of features extracted from first datasets PD1 Xsec and second datasets PD2 Xsec2024PF01558 12

[0067] recorded in cases in which the target 30 was either skin or not skin. Depicted are a second feature PSDiff 420, a third feature PDR 430, a fourth feature Amp 440, and a fifth feature Slope 450.

[0068] Fig. 7 shows a schematic depiction of a first average 451 and a first variance 452 of the fifth feature Slope 450 in the cases in which the target 30 is skin and a second average 453 and a second variance 454 of the fifth feature Slope 450 in the cases in which the target 30 is not skin. Fig. 7 also shows a first average 411 and a first variance 412 of a first feature FDSTD 410 in cases in which the target 30 is skin and a second average 413 and a second variance 414 of the first feature FDSTD 410 in cases in which the target 30 is not skin .

[0069] Figs . 6 and 7 show that each of the first feature FDSTD 410, the second feature PSDiff 420, the third feature PDR 430, the fourth feature Amp 440, and the fifth feature Slope 450 can be used to support the classification of the state of the sensor device 20.

[0070] For calculating the first feature FDSTD 410, a first derivative FD_PDl_Xsec of the first dataset PDl_Xsec and a first derivative FD_PD2_Xsec of the second dataset PD2_Xsec are calculated. Then, a variance of the derivative FD_PDl_Xsec of the first dataset PDl_Xsec is calculated as a first variance Var_FD_PDl and a variance of the first derivative FD_PD2_Xsec of the second dataset PD2_Xsec is calculated as a second variance Var_FD_PD2 . An absolute difference A_FD_Diff between the first variance Var_FD_PDl and the second variance Var_FD_PD2 and an absolute sum A_FD_Sum of the first variance Var_FD_PDl and the second variance Var_FD_PD2 are calculated:

[0071] A_FD_Diff = abs (Var_FD_PDl - Var_FD_PD2 )

[0072] A_FD_Sum = abs (Var_FD_PDl + Var_FD_PD2 ) ,

[0073] where abs (x) is the absolute value of X .2024PF01558 13

[0074] The first feature FDSTD 410 is computed as the ratio between the absolute difference A_FD_Diff and the absolute sum

[0075] AF FD Sum:

[0076] FDSTD = A_FD_Diff / A_FD_Sum.

[0077] For calculating the second feature PSDiff 420 a dataset M_PD_Xsec is calculated as the average of the first dataset PDl_Xsec and the second dataset PD2_Xsec such that the aver-age is calculated independently for each pair of correspond-ing first values PD1 and second values PD2 :

[0078] M_PD_Xsec =(PDl_Xsec + PD2_Xsec) / 2.

[0079] The dataset M_PD_Xsec of averages comprises the same length N as the first dataset PDl_Xsec and the second dataset PD2_Xsec . Then, an average value M_PD of the values of the dataset M_PD_Xsec is computed:

[0080] M_PD = mean (M_PD_Xsec) .

[0081] Alternatively, the average value M_PD could be calculated directly from the first dataset PDl_Xsec and the second dataset PD2_Xsec . Finally, the second feature PSDiff 420 is computed as

[0082] PSDiff = median ( abs ( PDl_Xsec - PD2_Xsec) / M_PD) ,

[0083] where median (X) is the median value of x and the vector notation means that the calculation is carried out for all elements of the vectors .

[0084] For calculating the third feature PDR 430, an offset second dataset PD2_Xsec_Of f set is calculated by adding an offset value to each element of the second dataset PD2 Xsec:

[0085] PD2 Xsec Offset PD2 Xsec + Offset .2024PF01558 14

[0086] In some variants, the offset value is a large number such as 1000 .

[0087] Next, a ratio dataset P_Xsec is calculated by dividing the elements of the first dataset PDl_Xsec by the corresponding elements of the offset second dataset PD2_Xsec_Of f set :

[0088] P_Xsec = PDl_Xsec / PD2_Xsec_Of f set .

[0089] The third feature PDR 430 is calculated as the average of the ratio dataset P_Xsec :

[0090] PDR = mean (P_Xsec) .

[0091] The fourth feature Amp 440 is calculated by computing a first central amplitude Ampl as a ratio between the central value (at index N div 2 ) of the first dataset PDl_Xsec and the average value M_PD:

[0092] Ampl = PDl_Xsec (N div 2 ) / M_PD.

[0093] Accordingly, a second central amplitude Amp2 is calculated as the ratio between a central value of the second dataset PD2_Xsec and the average value M_PD:

[0094] Amp2 = PD2_Xsec (N div 2 ) / M_PD.

[0095] Finally, the fourth feature Amp 440 is calculated as the geometric average of the first central amplitude Ampl and the second central amplitude Amp2 :

[0096] Amp = sqrt (Ampl * Amp2 ) ,

[0097] where sqrt (X) denotes the square root and the asterisk (* ) denotes multiplication.2024PF01558 15

[0098] The fifth feature Slope 450 is calculated by calculating a first Slopel as a ratio of the difference between the last value and the first value of the first dataset PDl_Xsec and the average value M_PD:

[0099] Slopel = (PDl_Xsec (N) - PDl_Xsec ( l ) ) / M_PD.

[0100] Accordingly, a second Slope2 is calculated as a ratio between the difference between the last value and the first value of the second dataset PD2_Xsec and the average value M_PD:

[0101] Slope2 = (PD2_Xsec (N) - PD2_Xsec ( l ) ) / M_PD.

[0102] Then, the fifth feature Slope 450 is calculated as the geometric average of the first slope Slopel and the second slope Slope2 :

[0103] Slope = KI * sqrt (Slopel * Slope2 ) ,

[0104] where KI is a first scaling factor .

[0105] A sixth feature SlopeCS that is not shown in Figs . 6 and 7, can be calculated by calculating a first cumulative dataset PDl_Xsec_CS from the values of the first dataset PDl_Xsec and a second cumulative dataset PD2_Xsec_CS from the values of the second dataset PD2_Xsec :

[0106] PDl_Xsec_CS = cumsum (PDl_Xsec)

[0107] PD2_Xsec_CS = cumsum (PD2_Xsec) ,

[0108] where cumsum (x) is the cumulative sum of the array x. A first cumulative slope SlopeCSl is calculated as a difference between the last and the first value of the first cumulative dataset PDl_Xsec_CS and a second cumulative slope SlopeCS2 is accordingly calculated as the difference between the last and the first value of the second cumulative dataset PD2 Xsec CS :

[0109] SlopeCSl = PDl_Xsec_CS (N) - PDl_Xsec_CS ( 1 )2024PF01558 16

[0110] SlopeCS2 = PD2_Xsec_CS (N) - PD2_Xsec_CS ( 1 ) .

[0111] Finally, the sixth feature SlopeCS is calculated as the geometric average between the first cumulative SlopeCSl and the second cumulative slope SlopeCS2 :

[0112] SlopeCS = K2 * sqrt (SlopeCSl * SlopeCS2 ) ,

[0113] where K2 is a second scaling factor .

[0114] It is possible to derive further features from the first dataset PDl_Xsec and the second dataset PD2_Xsec . It is also possible to combine several of the features described above into new features .

[0115] In one variant, the state of the sensor device 20 is classified as not facing human skin if the first feature FDSTD 410 is smaller than or equal to a pre-defined first state threshold value Tl . In this variant, the state of the sensor device 20 is classified as facing human skin if the first feature FDSTD 410 is larger than or equal to a pre-defined second state threshold value T2 . The state of the sensor device 20 is classified as indeterminate if the first feature FDSTD 410 is between the first state threshold value Tl and the second state threshold value T2 .

[0116] In other variants, other features such as the second feature PSDiff 420, the third feature PDR 430, the fourth feature Amp 440, the fifth feature Slope 450, and the sixth feature SlopeCS are used in addition or instead of the first feature FDSTD 410 for classifying the state of the sensor device 20 by comparing one or more of these features with one or more pre-defined threshold values . It is possible to combine the state predictions from several such comparisons into a final prediction for the classification of the state of the sensor device 20.2024PF01558 17

[0117] In some variants, the state of the sensor device 20 is classified in dependence of the first dataset PDl_Xsec and the second dataset PD2_Xsec using a pre-trained machine learning model such as a pre-trained neural network. In some variants, the machine learning model uses one or more of the features described above .

[0118] In some variants, it may be sufficient to record only the array of first values PD1 as the first dataset PDl_Xsec and classify the state of the sensor device 20 in dependence of only the first dataset PDl_Xsec . In this case, the average value M_PD can be calculated as the average of the first dataset PDl_Xsec and the first central amplitude Ampl or the first slope Slopel may be used as a feature for classifying the state of the sensor device, for example . Using only the first dataset PDl_Xsec may allow power savings and may provide an early classification of the state of the sensor device 20. In case that a further classification is needed, the second dataset PD2_Xsec of second values PD2 may be recorded afterwards .

[0119] The invention has been illustrated and described in more detail with the aid of exemplary variants . The invention is not, however, restricted to the examples disclosed. Rather, other variants may be derived therefrom by the person skilled in the art .2024PF01558 18

[0120] REFERENCE SYMBOLS

[0121] 10 electronic device

[0122] 11 earbud

[0123] 20 sensor device

[0124] 21 detection side

[0125] 30 target

[0126] 100 emission unit

[0127] 110 first emitter

[0128] 111 first light ( first wavelength)

[0129] 120 second emitter

[0130] 121 second light (second wavelength)

[0131] 200 detection unit

[0132] 210 first detector

[0133] 211 first intensity ( first wavelength) 220 second detector

[0134] 221 second intensity (second wavelength)

[0135] 300 time

[0136] 410 first feature FDSTD

[0137] 411 first average

[0138] 412 first variance

[0139] 413 second average

[0140] 414 second variance

[0141] 420 second feature PSDiff

[0142] 430 third feature PDR

[0143] 440 fourth feature Amp

[0144] 450 fifth feature Slope

[0145] 451 first average

[0146] 452 first variance

[0147] 453 second average

[0148] 454 second variance2024PF01558 19

[0149] PDl_Xsec first dataset

[0150] PD1 first value

[0151] PD2_Xsec second dataset

[0152] PD2 second value

[0153] N length of dataset

[0154] FD_PDl_Xsec first derivative of the first dataset FD_PD2_Xsec first derivative of the second dataset Var_FD_PDl first variance

[0155] Var_FD_PD2 second variance

[0156] A_FD_Dif f absolute difference

[0157] A_FD_Sum absolute sum

[0158] FDSTD first feature

[0159] T1 first state threshold value

[0160] T2 second state threshold value

[0161] M_PD_Xsec dataset of averages

[0162] M_PD average value

[0163] PSDif f second feature

[0164] PD2_Xsec_0f f set offset second dataset

[0165] P_X sec ratio dataset

[0166] PDR third feature

[0167] Ampl first central amplitude

[0168] Amp 2 second central amplitude

[0169] Amp fourth feature

[0170] Slopel first slope

[0171] Slope2 second slope

[0172] KI first scaling factor

[0173] Slope fifth feature

[0174] PDl_Xsec_CS first cumulative dataset PD2_Xsec_CS second cumulative dataset

[0175] SlopeCSl first cumulative slope

[0176] SlopeCS2 second cumulative slope2024PF01558 - 20 -

[0177] K2 second scaling factor SlopeCS sixth feature

Claims

2024PF01558 21CLAIMS1 . A method for operating a sensor devicecomprising- recording an array of first values as a first dataset and an array of second values as a second dataset, wherein recording a first value includes- emitting light having a first wavelength;- detecting an intensity of reflected light having the first wavelength as the first value;and recording a second value includes- emitting light having a second wavelength;- detecting an intensity of reflected light having the second wavelength as the second value;- classifying a state of the sensor device in dependence of the first dataset and the second dataset .

2. The method as claimed in claim 1,wherein classifying the state comprises a classification of whether the sensor device is facing skin.

3. The method as claimed in one of the preceding claims, wherein recording the array of first values and the array of second values is started when the first values exceed a first intensity threshold value or the second values exceed a second intensity threshold value .

4. The method as claimed in one of the preceding claims, wherein the array of first values and the array of second values is recorded over a pre-defined period of time between 0.5 seconds and 5 seconds, in particular a pre-defined period of time between 1 second and 3 seconds .

5. The method as claimed in one of the preceding claims, wherein the state of the sensor device is classified in dependence of a temporal development of the first values and the second values .2024PF01558 226. The method as claimed in one of the preceding claims, wherein classifying the state of the sensor device includes- filtering the array of first values and the array of second values using a zero-phase linear filter .

7. The method as claimed in one of the preceding claims, wherein classifying the state of the sensor device includes- calculating a first derivative of the first dataset and a first derivative of the second dataset,- calculating a variance of the first derivative of the first dataset as a first variance and calculating a variance of the first derivative of the second dataset as a second variance;- calculating an absolute difference between the first variance and the second variance;- calculating an absolute sum of the first variance and the second variance;- calculating a feature as the ratio between the absolute difference and the absolute sum;wherein the state of the sensor device is classified in dependence of the feature .

8. The method as claimed in claim 7,wherein the state of the sensor device is classified as not facing human skin if the feature is smaller than or equal to a first state threshold value,wherein the state of the sensor device is classified as facing human skin if the feature is larger than or equal to a second state threshold value,wherein the state of the sensor device is classified as indeterminate if the feature is between the first state threshold value and the second state threshold value .

9. A method for operating a sensor devicecomprising- recording an array of first values as a first dataset,2024PF01558 23wherein recording a first value includes- emitting light having a first wavelength;- detecting an intensity of reflected light having the first wavelength as the first value;- classifying a state of the sensor device in dependence of the first dataset .

10. The method as claimed in claim 9,wherein classifying the state comprises a classification of whether the sensor device is facing skin.

11. The method as claimed in one of claims 9 and 10, wherein recording the array of first values is started when the first values exceed a first intensity threshold value .

12. The method as claimed in one of claims 9 to 11, wherein classifying the state of the sensor device includes- filtering the array of first values using a zero-phase linear filter .

13. A sensor devicecomprisingan emission unit for emitting light having a first wavelength and emitting light having a second wavelength, a detection unit for detecting an intensity of reflected light having the first wavelength and detecting an intensity of reflected light having the second wavelength, wherein the sensor device is configured to carry out a method as claimed in one of claims 1 to 12.

14. An electronic devicecomprising a sensor device as claimed in claim 13.

15. The electronic device as claimed in claim 14,wherein the electronic device is an earbud.

Images