US20230148445A9 - Method for identifying occupant-specific settings and vehicle for performing method for identifying occupant-specific settings - Google Patents
Method for identifying occupant-specific settings and vehicle for performing method for identifying occupant-specific settings Download PDFInfo
- Publication number
- US20230148445A9 US20230148445A9 US16/697,739 US201916697739A US2023148445A9 US 20230148445 A9 US20230148445 A9 US 20230148445A9 US 201916697739 A US201916697739 A US 201916697739A US 2023148445 A9 US2023148445 A9 US 2023148445A9
- Authority
- US
- United States
- Prior art keywords
- vehicle
- cluster
- occupant
- descriptor
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
- B60R16/00—Electric or fluid circuits specially adapted for vehicles and not otherwise provided for; Arrangement of elements of electric or fluid circuits specially adapted for vehicles and not otherwise provided for
- B60R16/02—Electric or fluid circuits specially adapted for vehicles and not otherwise provided for; Arrangement of elements of electric or fluid circuits specially adapted for vehicles and not otherwise provided for electric constitutive elements
- B60R16/037—Electric or fluid circuits specially adapted for vehicles and not otherwise provided for; Arrangement of elements of electric or fluid circuits specially adapted for vehicles and not otherwise provided for electric constitutive elements for occupant comfort, e.g. for automatic adjustment of appliances according to personal settings, e.g. seats, mirrors, steering wheel
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/59—Context or environment of the image inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/117—Identification of persons
- A61B5/1171—Identification of persons based on the shapes or appearances of their bodies or parts thereof
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/117—Identification of persons
- A61B5/1171—Identification of persons based on the shapes or appearances of their bodies or parts thereof
- A61B5/1172—Identification of persons based on the shapes or appearances of their bodies or parts thereof using fingerprinting
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/117—Identification of persons
- A61B5/1171—Identification of persons based on the shapes or appearances of their bodies or parts thereof
- A61B5/1176—Recognition of faces
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- G06K9/00006—
-
- G06K9/00221—
-
- G06K9/00597—
-
- G06K9/00892—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/18—Eye characteristics, e.g. of the iris
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/70—Multimodal biometrics, e.g. combining information from different biometric modalities
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification
-
- G10L17/005—
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification
- G10L17/04—Training, enrolment or model building
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Molecular Biology (AREA)
- Veterinary Medicine (AREA)
- Heart & Thoracic Surgery (AREA)
- Biomedical Technology (AREA)
- Pathology (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Acoustics & Sound (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Mechanical Engineering (AREA)
- Software Systems (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Ophthalmology & Optometry (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Collating Specific Patterns (AREA)
Abstract
Description
- The present invention relates to a method for identifying occupant-specific settings for a vehicle and a vehicle capable of performing such method. In particular, the present invention relates to a method for identifying occupant-specific settings in an unsupervised manner and a vehicle capable of performing such method.
- Some vehicles are provided with the capability of adjusting particular vehicle settings, such as a seat position, steering wheel position, mirror position, etc., based on a stored setting. Such stored setting can be based on a user identity, so that multiple occupants (driver and/or passenger) can customize particular vehicle items.
- The identification of vehicle occupants is conventionally determined on the basis of an unambiguous identity of the occupant. For instance, the identity of the occupant is determined based on a pre-registered device, such as a key, a token or mobile telephone, or based on a pre-registered user profile identifiable by a name, password, etc., or based on inalterable biometric data, such as a fingerprint, iris, speech, etc. Such an identification system requires a supervised learning phase, during which an occupant account is created and stored in a vehicle system.
- It is an object of the present invention to provide a more efficient way of identifying occupant-specific settings for a vehicle.
- This object is solved by a method as defined in independent claim 1, and a vehicle as defined in
independent claim 12. Preferred embodiments are defined by the dependent claims. - A method for identifying occupant-specific settings for a vehicle comprises the steps of detecting a vehicle occupant, such as a driver or passenger, and collecting biometric occupant data of the detected vehicle occupant. The detecting of a vehicle occupant is directed to the detecting of a presence of the vehicle occupant, but does not include detecting an identification of a particular vehicle occupant. For instance, the method does not require detecting a particular vehicle occupant on the basis of unique occupant data, such as a key or (RFID-) token.
- The method further comprises (once the presence of a vehicle occupant is detected and/or another event has triggered the collection of biometric occupant data) assembling a descriptor based on at least part of the collected occupant data, storing the descriptor in a database, and updating a cluster of descriptors in the database based on the assembled descriptor. A descriptor stores at least one feature, such as a biometric feature, numerically, and in case of multiple features of the vehicle occupant, the descriptor stores and combines the different features numerically as a vector. The descriptors grouped in the cluster are close to each other, e.g. have a distance below a predefined threshold, and the cluster corresponds to the detected vehicle occupant. In other words, the database holds (stores) at least two descriptors that can be compared with one another. If two or more descriptors are close to each other these descriptors can be grouped together in a cluster representing the detected vehicle occupant.
- Thus, the database can further store a plurality of clusters, i.e. a plurality of descriptors grouped in two or more clusters representing two or more vehicle occupants, respectively. This provides for identifying vehicle occupants in an unsupervised manner, since there is no requirement for inputting personal information in a preceding user profile setup. New or unknown vehicle occupants can be identified “on-the-fly”, i.e. without initial user profile setup, simply by supplementing an additional cluster in the database starting with a descriptor having a distance exceeding the predefined threshold to each descriptor already in the database. This allows for vehicle occupant identification without distracting the occupant, in particular a driver of the vehicle. In addition, the biometric data of the vehicle occupant may be of a rather general type. For instance, it is not required that the vehicle occupant operates a particular sensor device, such as a fingerprint sensor, and iris sensor, etc. due to the collection of different features of the vehicle occupant and storing multiple descriptors of the same vehicle occupant in a cluster. This further allows identification of the vehicle occupant without a single unambiguous or unique biometric feature.
- The numerical interpretation of the biometric and/or other feature(s) of and relating to the vehicle occupant may include the storing of sensor value data, such as weight, height (relative to a fixed object in the vehicle, such as a back rest of the seat) and safety belt length, and/or the storing of calculated data based on sensor data, such as an average hair colour, skin colour, colour of the iris or the like, the curvature of an outline of the occupant's torso, specific features of the occupant's voice, etc. In addition or alternatively, a descriptor may include data based on captured three-dimensional (3D) images, such as a 3D geometry and/or topography of a head, face, extremity, torso, etc. of the vehicle occupant. For instance, the descriptor can include data representing a distance of the eyes, distance(s) of eye(s) and nose and/or ears, a size of an ear, volume of the head and/or an extremity, etc. It is to be understood that the descriptor may further store biometric data of or derived from a fingerprint sensor and/or an iris sensor.
- Additionally and/or alternatively, a descriptor may include data-driven features like embeddings calculated by artificial neural networks and/or related machine learning methods applied on (sensor) data representing the vehicle occupant's face, body, speech, etc.
- In an implementation variant the updating of a cluster of descriptors may comprise calculating a distance between the assembled descriptor and at least one further descriptor stored in the database, and associating the assembled descriptor with a cluster defined by the at least one further descriptor, if the distance falls below a predefined threshold or a threshold evaluated (learned) over time by the system. In other words, a cluster may already exist in the database that groups the at least one further descriptor, so that each of the further descriptors of the cluster can be used to identify the cluster. In the descriptor space distances between descriptors can be computed using any suitable distance measure, for example, calculating a Euclidean distance, a geodesic distance, etc. Thus, occupants of the vehicle may be identified by analysing distances of the descriptor(s) representing a specific occupant to descriptor(s) representing another occupant, for example by finding best matches with minimum distances in the descriptor space. The distance between descriptors may be calculated on a feature basis constituting the descriptor. In other words, a distance between descriptors is calculated on the basis of distances between corresponding features in the descriptors. Thus, a vehicle occupant may still be identified, even if the feature changes, such as hair colour.
- In addition or alternatively, a distance between the assembled descriptor and an averaged descriptor representing an average of all descriptors of the cluster may be calculated and compared with the predefined threshold. This avoids increasing the “size” of the cluster, i.e. avoids a distance between two descriptors of the same cluster that exceeds a multiple of the predefined distance. This reduces the calculation of false positives, i.e. identification of an alleged vehicle occupant due to an oversized cluster, due to a deviation of the descriptors in one cluster from a mean value (average) being too large.
- In the present disclosure reference is made to a distance between descriptors and/or between a descriptor and a mean (average) value of the cluster. It is to be understood that instead of or in addition to a distance the method may also rely on a similarity measure (the similarity of descriptors) and/or probability measure (a probability that two descriptors represent the same vehicle occupant). Thus, a close similarity and a high probability would be employed as is a small distance.
- Furthermore, associating the assembled descriptor with the cluster may comprise storing a cluster identification for or with the descriptor (for example supplementing the descriptor with a cluster identification) and/or storing descriptor identifications of all descriptors constituting the cluster, such as storing a mapping table of descriptors and clusters.
- In another implementation variant the biometric occupant data comprises strong biometric data and/or soft biometric data. A differentiation between strong and soft biometric data allows reliable identification of a particular vehicle occupant, even if a change of the soft biometric data of the same occupant is monitored, due to storing such changes in the cluster representing the occupant over time. For instance, strong biometric data may include facial feature data (e.g., geometric features of a face of the occupant), speech data, iris data and/or fingerprint data. The strong biometric data can be unique to the occupant, such as iris and fingerprint data. The soft biometric data may include body weight data, body height data and/or safety belt length data. The safety belt length data may depend on a body height of the occupant and, hence, represents soft biometric data, which may change or vary over time.
- The method is not relying only on strong biometric data, since vehicle occupant identification is based on the entire descriptor. Even if a particular biometric data item may not be derived, a descriptor can still be stored and clustered, since it provides an identification of the vehicle occupant. Due to the multiplicity of descriptors clustered to identify a vehicle occupant a reliable identification is possible, even without strong biometric data. According to a further variant, a descriptor may be discarded, if its distance to other descriptors in the database exceeds the (predefined or learned) threshold and/or the number of features or types of biometric data available at the time the descriptor is assembled and/or stored falls below another threshold.
- Thus, instead of requiring a unique identification (such as a token or key) an occupant of the vehicle may be identified based on a variety of features describing the occupant. Furthermore, an identity of the occupant, such as a name, age, sex, etc. is not required which avoids problems of data protection. In other words, although the vehicle occupant can be identified, the information is anonymized as it does not comprise any personal information about the vehicle occupant.
- The data constituting a descriptor may include all collected biometric occupant data or, alternatively, may include a subset of the collected biometric occupant data. For instance, based on a predefined setting of the method, only soft biometric data may be used to assemble a descriptor, only strong biometric data may be used to assemble a descriptor, only some features of the strong and/or soft biometric data may be used to assemble a descriptor. In other words, a vehicle occupant is modelled by a descriptor numerically by combining different features as a vector. These features may be based on real sensor measurement data or abstract mathematical descriptions thereof. The combination of multiple features, i.e. the use of multiple biometric occupant data, to assemble a descriptor provides for more discriminative descriptors of the vehicle occupant.
- According to a variant, at least the steps of collecting biometric occupant data, assembling a descriptor, storing the descriptor, and updating the cluster of descriptors may be performed automatically without user initiation and/or user intervention. For instance, performing these steps can be triggered by a system of the vehicle at particular time points, such as the beginning of a journey, the start of an engine of the vehicle, in certain time intervals (e.g., once per hour, once a day, once a week, etc.), and/or when a state of the vehicle changes, such as when a door is opened and/or closed, when a safety belt is buckled, and/or when a signal of a weight sensor in a seat of the vehicle changes and/or indicates a change of weight above a (predefined) threshold.
- In a further implementation variant a cluster algorithm for clustering the assembled descriptors can be selected. For instance, a cluster algorithm can be chosen on the basis of pre-set task settings. According to a first setting, if the number of occupants and, hence, the number of clusters is known or predictable, a centroid-based approach is used, such as a k-means approach. According to a second setting, if the number of vehicle occupants and, hence, the number of clusters are unknown and arbitrary large, for example in a vehicle of a large fleet used by a plurality of drivers, a hierarchical clustering is used.
- In yet a further implementation variant, the clustering (cluster algorithm) is triggered by certain events. For instance, after collecting/assembling a predefined number of descriptors, the clustering is updated. The predefined number may be 1, but in an alternative implementation may be greater than 1, such as after collecting/assembling 5 descriptors or after collecting/assembling 10 descriptors. In addition or alternatively, the clustering may be updated after the vehicle occupant adjusted a particular vehicle setting, such as adjusting a seat position, adjusting a mirror, or the like.
- According to another variant the method may further comprise associating the descriptor with at least one current vehicle setting. For instance, storing the descriptor may comprise storing the descriptor in association with the at least one vehicle setting. The storing may take place in the same database where the descriptor(s) and clusters are stored. A vehicle setting may relate to any item in the vehicle that can be adjusted and/or personalized, such as the position of a seat, the position of a back rest, the position of a head rest, the position of a mirror, the position of a steering wheel, the position of a safety belt holder or deflector, the temperature and/or volume flow of an air conditioning system, a radio and/or TV station or other setting of an infotainment system, etc.
- According to another implementation variant, the method may further comprise calculating a distance of the assembled descriptor to each cluster in the database, and comparing the calculated distance with a threshold value, such as a predefined threshold. The distance between the descriptor and cluster may be based on an average for each feature of the descriptors constituting the cluster. In other words, the assembled descriptor is compared to an average description of the vehicle occupant represented by the cluster. The distance may be calculated on a feature-by-feature basis.
- If the calculated distance exceeds the threshold value, it is assumed that the detected occupant of the vehicle is unknown to the system. In this case, the method comprises storing the assembled descriptor in association with a new cluster representing a new vehicle occupant. Alternatively, the assembled descriptor may be stored in association with a cluster for unknown descriptors, which will be evaluated at a later time point, such as a time when two unknown clusters having a distance below the threshold value are assembled and stored. In any case, a learning phase of the system is initiated, in which a new occupant is described via descriptors and vehicle settings are stored for the new occupant. Alternatively, the descriptor may be discarded, for example, if the number of features (number of biometric data) available for this respective descriptor falls below a predefined threshold.
- Furthermore, the method also comprises, if the calculated distance does not exceed the threshold value, selecting the best matching cluster stored in the database, and calculating a vehicle setting based on the best matching cluster. In other words, the best matching cluster represents the vehicle occupant that has been identified via the assembled descriptor, i.e. that has been identified as a particular occupant on a particular seat of the vehicle.
- Once calculated, the calculated vehicle setting can be used to adjust at least one vehicle item. Thus, once a particular occupant has been detected and identified via the assembled descriptor and the clusters stored in the database (in form of a matching cluster and not a specific identification as in conventional systems), a vehicle item can automatically be adjusted on the basis of vehicle setting information stored for the best matching cluster, i.e. for the particular occupant.
- Calculating a vehicle setting may be based on any function deriving an optimum vehicle setting for the particular occupant. According to a variant, calculating a vehicle setting comprises retrieving at least one cluster item of the best matching cluster from the database, i.e. at least one cluster item stored in association with the best matching cluster in the database. Each cluster item may include at least one vehicle setting. The calculating of a vehicle setting may further comprise calculating at least one occupant-specific vehicle setting based on the at least one vehicle setting of the retrieved at least one cluster item. In other words, a cluster item associated with the particular cluster may be a feature vector representing the setting of one or more vehicle items in association to the particular cluster. The association to a particular cluster specifies an occupant-specific vehicle setting of the particular vehicle item identified by the cluster item.
- According to a further variant, the calculating of at least one occupant-specific vehicle setting comprises determining a latest stored vehicle setting, such as a last tuned radio and/or TV station, calculating an average or median or mode of a predefined (or predetermined) number of vehicle settings of the retrieved at least one cluster item, such as an average seat position or median or mode of seat positions, an average mirror position or median or mode of mirror positions, an average temperature setting or median or mode of temperature settings, etc., and/or determining a vehicle setting based on sensor data, such as selecting an air conditioning setting out of a plurality of stored air conditioning settings (from stored cluster items) depending on a current temperature measurement (in this case the cluster items have to be stored together with a temperature measurement when the particular air conditioning setting was set and stored). The mode of vehicle settings is a particular setting having the highest number of occurrences in the retrieved at least one cluster item.
- The cluster item(s) may be stored each time a descriptor is assembled and/or a descriptor is stored in the database. Thus, not only the vehicle occupant is monitored over time, in order to develop a cluster of descriptors representing the vehicle occupant, but also the vehicle setting(s) is/are monitored over time. As with the descriptors representing the vehicle occupant, the vehicle setting(s) may be stored in a descriptor, which is clustered and associated with the cluster representing the vehicle occupant. Thus, vehicle item(s) can be set not only in accordance with the particular vehicle occupant, but also in accordance with further influencing factors, such as outside temperature, rain, sunshine, speed of travel, type of road currently being used, etc.
- According to a further aspect, a computer-readable medium is configured to store executable instructions that, when executed by a processor, cause the processor to perform the method according to one of the disclosed (implementation) variants. The computer-readable medium can be any volatile or non-volatile storage medium.
- According to yet a further aspect, a vehicle comprises a processor, a storage device including a database, and at least one adjustable vehicle item configured to assume a particular setting. The processor is configured to perform the method according to one of the disclosed (implementation) variants.
- The vehicle may further comprise at least one sensor configured to sense the biometric occupant data. The at least one sensor may optionally comprise a camera, a microphone, a weight sensor, and/or a fingerprint sensor. The vehicle may additionally comprise at least one sensor configured to sense an ambient feature of the vehicle, such as the temperature sensor, a water sensor, a vibration sensor, a gyroscope, etc. Such ambient feature detected by the additional sensor(s) can be used by the processor to calculate a vehicle setting based on ambient influencing factors.
- In an implementation variant the processor is further configured to initiate the method according to one of the disclosed (implementation) variants each time the vehicle starts a journey.
- The present disclosure is not restricted to the described aspects, variants, implementation variants and alternatives, but combinations of the described aspects, variants, implementation variants and alternatives our possible and fall under the present disclosure.
- Preferred examples for understanding the disclosed method for identifying occupant-specific settings for a vehicle and corresponding vehicle are described in greater detail with reference to the attached schematic drawings in the following, wherein
-
FIG. 1 depicts a schematic flow diagram of an inference phase for identifying a best matching cluster representing a detected vehicle occupant, -
FIG. 2 depicts a schematic flow diagram of a learning phase for assembling and storing a descriptor describing the detected vehicle occupant, -
FIG. 3 depicts a schematic flow diagram of the retrieval phase for retrieving occupant-specific settings for a vehicle, -
FIG. 4 schematically illustrates a vehicle comprising a system capable of performing a method according to the phases ofFIGS. 1 to 3 , and -
FIG. 5 schematically illustrates a descriptor used for describing the detected vehicle occupant. -
FIG. 1 depicts a schematic flow diagram of aninference phase 100 for identifying a best matching cluster representing a detected vehicle occupant. Theinference phase 100 assumes that adatabase 10 already holds vehicle occupant-specific data, such asdata 11 representing the occupant anddata 12 representing vehicle settings. In step 101 a vehicle occupant is detected, for example due to the change of a sensor signal, such as a camera 41 or microphone 41 and/or a weight sensor 42 (seeFIG. 4 ). - Once the vehicle occupant is detected (or alternatively a different event triggers the
inference phase 100, such as the change of an ambient feature detected by an ambient sensor (not shown)), biometric data of the detected vehicle occupant is collected (step 110). The collection of this data may include strong and/or soft biometric occupant data, which is further used to assemble adescriptor 11 instep 120. Assembling adescriptor 11 can include the calculation of feature values to be stored as descriptor items as will be described with respect toFIG. 5 . - The method then calculates (130) a distance “d” of the assembled
descriptor 11 toclusters 13 retrieved fromdatabase 10. Theclusters 13 are a group ofdescriptors 11 having a distance to one another falling below a predefined threshold, so that acluster 13 represents a vehicle occupant. Instep 140 it is determined whether the calculated distance ofdescriptor 11 to cluster(s) 13 exceeds a predefined threshold do or not. - If the distance of the
descriptor 11 to aparticular cluster 13 does not exceed the threshold value, abest matching cluster 13 is selected instep 150. The selection of thebest matching cluster 13 may include selecting acluster 13 having the smallest distance to the assembleddescriptor 11. In other words, thebest matching cluster 13 represents a vehicle occupant that has been identified on the basis ofdescriptor 11 data andcluster 13 data stored indatabase 10. Based on the best matching cluster 13 aretrieval phase 300 is started. - If the distance “d” of the assembled
descriptor 11 to eachcluster 13 stored indatabase 10 exceeds the predefined threshold d0, alearning phase 200 schematically illustrated inFIG. 2 is started. Alternatively, thelearning phase 200 may be started when detecting 101 a vehicle occupant, i.e. may be started due to the same trigger initiating theinference phase 100. Furthermore, the detecting 101 of a vehicle occupant may be based on the same evaluations as described with respect to theinference phase 100. - Likewise, the collecting 110 of biometric occupant data and the assembling 120 of a
descriptor 11, may correspond to the steps of theinference phase 100. For instance, when thelearning phase 200 is initiated fromstep 140 of theinference phase 100, the method may directly start (at point “A”) withstep 210, i.e. associating the assembleddescriptor 11 with one ormore car settings 12. The associating 210 of the assembleddescriptor 11 with one ormore car settings 12 may be based on thecar settings 12 as set at the time of assembling thedescriptor 11 or storing 220 thedescriptor 11 indatabase 10. - Alternatively, the associating 210 of the
descriptor 11 withcar settings 12 may be skipped and the method (starts with or) proceeds withstep 220, i.e. storing (only) thedescriptor 11 indatabase 10. This may be useful, if a new (empty)database 10 not holding any ormany descriptors 11, so that thedatabase 10 is first filled withdescriptors 11 describing one or more vehicle occupants, beforepersonalized car settings 12 are stored for the vehicle occupants represented by the storeddescriptors 11. - For instance, in step 230 a determination may be made whether the
learning phase 200 shall be continued or whether the stored (220)descriptors 11 shall be further evaluated and/or analysed. Thus, the method may repeat beginning atstep 110, i.e. collecting biometric occupant data. Otherwise, the method may continue withstep 240, i.e. updating clusters 13 (or clustering) of the assembled and storeddescriptors 11. Acluster 13 is used togroup descriptors 11, which are close to each other, i.e. that have a distance between one another falling below a predefined threshold (such as threshold d0). Alternatively,descriptors 11 having a distance to an average or median or mode of thecluster 13 falling below the predefined threshold (e.g., d0) are grouped in thecluster 13. Thus, acluster 13 corresponds to the detected vehicle occupant and describes features of the vehicle occupant. -
FIG. 3 depicts a schematic flow diagram of theretrieval phase 300 for retrieving occupant-specific settings 15 for a vehicle 1. Theretrieval phase 300 starts with abest matching cluster 13, such as the one selected in step 150 (FIG. 1 ). Based on thebest matching cluster 13 at least one cluster item is retrieved fromdatabase 10 instep 310, wherein each cluster item includes at least one vehicle setting 12. - Thereafter, a (desired) vehicle setting is calculated 320 from the retrieved cluster item(s), i.e. from the retrieved setting(s). The calculating 320 of at least one occupant-specific vehicle setting 15 can comprise determining the latest stored vehicle setting 12, calculating an average of a predefined number of
vehicle settings 12 of the retrieved at least one cluster item, and/or determining a vehicle setting 12 based on sensor data, such as data of sensors detecting ambient features. - Once the desired vehicle setting(s) is/are calculated (step 320), the corresponding vehicle item(s), such as a vehicle seat 60 (
FIG. 4 ), is/are adjusted instep 330 to the calculated vehicle setting(s). Theretrieval phase 300 is then terminated. -
FIG. 4 schematically illustrates a vehicle 1 comprising asystem 50 capable of performing a method according to one or more of theinference phase 100, learningphase 200, andretrieval phase 300. Thesystem 50 may be a processor and may further include adatabase 10 capable of storingdescriptors 11,vehicle settings 12 andclusters 13. Furthermore, thesystem 50 may also include a storage device, such as a computer-readable medium, configured to store executable instructions that, when executed by a processor perform one or more of thephases - The
system 50 may be part of the vehicle 1, for example may be integrated into an electronic control unit (ECU) forming part of the vehicle 1. - In addition, the vehicle may further include at least one
sensor 41, 42 configured to sense biometric occupant data.FIG. 4 exemplarily depicts a camera or microphone 41 and aweight sensor 42 installed in or on aseat 60 of the vehicle 1. Theseat 60 further illustrates an exemplary vehicle item that can be adjusted or personalized for the detected vehicle occupant. In other words, occupant-specific vehicle settings 15 may be retrieved in theretrieval phase 300 and may be used, for example byprocessor system 50 to adjusted a setting of thevehicle item 60. -
FIG. 5 schematically illustrates adescriptor 11 used for describing the detected vehicle occupant. The data stored in and/or with thedescriptor 11 may comprise (strong) biometric data (biometric attributes) 20 as well as soft biometric data (of biometric attributes) 30. The (strong)biometric data 20 can includefacial feature data 21,speech data 22, iris data and/or fingerprint data, while the softbiometric data 30 can include weight data 31 of a body of the occupant,body height data 32 and/or safety belt length data. - A data structure similar to the
descriptor 11 can be used for storing thevehicle settings 12. In other words, a similar descriptor can be stored indatabase 10 containing data and data values representing the setting of particular vehicle items, such as any item in the vehicle that can be adjusted or personalized for the respective occupant, in association with adescriptor 11 and/or in association with acluster 13.
Claims (15)
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP18211653.3 | 2018-12-11 | ||
EP18211653.3-1132 | 2018-12-11 | ||
EP18211653 | 2018-12-11 | ||
EP18211653.3A EP3666596B1 (en) | 2018-12-11 | 2018-12-11 | Method for identifying occupant-specific settings and vehicle for performing method for identifying occupant-specific settings |
Publications (3)
Publication Number | Publication Date |
---|---|
US20200180534A1 US20200180534A1 (en) | 2020-06-11 |
US20230148445A9 true US20230148445A9 (en) | 2023-05-11 |
US11679730B2 US11679730B2 (en) | 2023-06-20 |
Family
ID=64664885
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/697,739 Active 2041-03-24 US11679730B2 (en) | 2018-12-11 | 2019-11-27 | Method for identifying occupant-specific settings and vehicle for performing method for identifying occupant-specific settings |
Country Status (3)
Country | Link |
---|---|
US (1) | US11679730B2 (en) |
EP (1) | EP3666596B1 (en) |
CN (1) | CN111310551B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE102018206717A1 (en) * | 2018-05-02 | 2019-11-07 | Audi Ag | Method for the driving situation-dependent operation of a motor vehicle system of a motor vehicle, personalization device and motor vehicle |
US11238067B2 (en) * | 2018-07-17 | 2022-02-01 | Intel Corporation | Systems, methods and apparatus to cluster and classify individuals in groups |
US11613218B2 (en) | 2020-12-14 | 2023-03-28 | Ford Global Technologies, Llc | Seat adjustment |
JP2022185424A (en) * | 2021-06-02 | 2022-12-14 | マツダ株式会社 | Body-type class identification device and method and driving-posture setting device and method |
EP4300443A1 (en) * | 2022-06-29 | 2024-01-03 | Volkswagen Ag | Method for predicting a vehicle-specific parameter, method for creating a model, computer program product and system |
Family Cites Families (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6810309B2 (en) * | 2002-04-25 | 2004-10-26 | Visteon Global Technologies, Inc. | Vehicle personalization via biometric identification |
DE102005021171A1 (en) | 2005-05-06 | 2006-11-16 | Daimlerchrysler Ag | Method for setting an occupant protection and / or comfort system and occupant protection and / or comfort system |
DE102010002928B4 (en) * | 2010-03-16 | 2023-11-02 | Bayerische Motoren Werke Aktiengesellschaft | Automotive configuration system |
EP2797794A4 (en) * | 2011-12-29 | 2017-01-04 | Intel Corporation | Systems, methods, and apparatus for identifying an occupant of a vehicle |
WO2013133791A1 (en) * | 2012-03-05 | 2013-09-12 | Intel Corporation | User identification and personalized vehicle settings management system |
CN103065151B (en) * | 2012-11-04 | 2017-02-22 | 北京工业大学 | Vehicle identification method based on depth information |
CN103280108B (en) * | 2013-05-20 | 2015-04-22 | 中国人民解放军国防科学技术大学 | Passenger car safety pre-warning system based on visual perception and car networking |
CN103310233B (en) * | 2013-06-28 | 2016-03-23 | 青岛科技大学 | With similarity method for digging between class behavior multi views and Activity recognition method |
US20150046060A1 (en) * | 2013-08-12 | 2015-02-12 | Mitsubishi Electric Research Laboratories, Inc. | Method and System for Adjusting Vehicle Settings |
US9193359B2 (en) | 2013-08-12 | 2015-11-24 | GM Global Technology Operations LLC | Vehicle systems and methods for identifying a driver |
US9475496B2 (en) * | 2013-11-22 | 2016-10-25 | Ford Global Technologies, Llc | Modified autonomous vehicle settings |
DE102015207774A1 (en) * | 2015-04-28 | 2016-11-03 | Volkswagen Aktiengesellschaft | Device for identifying a driver and vehicle |
US10380557B2 (en) * | 2015-07-31 | 2019-08-13 | Snap-On Incorporated | Methods and systems for clustering of repair orders based on alternative repair indicators |
EP3159853B1 (en) * | 2015-10-23 | 2019-03-27 | Harman International Industries, Incorporated | Systems and methods for advanced driver assistance analytics |
CN105976140B (en) * | 2016-04-27 | 2019-10-11 | 大连海事大学 | Vehicle goods real-time matching method under extensive stream data environment |
DE102016008339A1 (en) * | 2016-06-21 | 2017-02-09 | Daimler Ag | Method for identifying a vehicle user in a vehicle |
US10850693B1 (en) * | 2018-04-05 | 2020-12-01 | Ambarella International Lp | Determining comfort settings in vehicles using computer vision |
US11040619B1 (en) * | 2018-04-05 | 2021-06-22 | Ambarella International Lp | Limiting car behavior based on a pre-set driver profile enabled by face recognition |
US10953850B1 (en) * | 2018-04-05 | 2021-03-23 | Ambarella International Lp | Seatbelt detection using computer vision |
US10867218B2 (en) * | 2018-04-26 | 2020-12-15 | Lear Corporation | Biometric sensor fusion to classify vehicle passenger state |
CN108734129A (en) * | 2018-05-21 | 2018-11-02 | 上海应用技术大学 | mobile phone and vehicle location analysis method and system |
US10949659B2 (en) * | 2019-04-16 | 2021-03-16 | GM Global Technology Operations LLC | Vehicle occupant detection |
US11371965B2 (en) * | 2019-07-10 | 2022-06-28 | Vibrant Corporation | Digital twin model inversion for testing |
-
2018
- 2018-12-11 EP EP18211653.3A patent/EP3666596B1/en active Active
-
2019
- 2019-11-27 US US16/697,739 patent/US11679730B2/en active Active
- 2019-12-11 CN CN201911266950.8A patent/CN111310551B/en active Active
Also Published As
Publication number | Publication date |
---|---|
US11679730B2 (en) | 2023-06-20 |
CN111310551A (en) | 2020-06-19 |
EP3666596A1 (en) | 2020-06-17 |
EP3666596B1 (en) | 2021-04-28 |
CN111310551B (en) | 2023-07-18 |
US20200180534A1 (en) | 2020-06-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11679730B2 (en) | Method for identifying occupant-specific settings and vehicle for performing method for identifying occupant-specific settings | |
US8238617B2 (en) | Vehicle operation control device and method, as well as, program | |
CN104816694B (en) | One kind is driven condition intelligent adjusting apparatus and method | |
US9573541B2 (en) | Systems, methods, and apparatus for identifying an occupant of a vehicle | |
US9663112B2 (en) | Adaptive driver identification fusion | |
US20200110952A1 (en) | System and method for determining probability that a vehicle driver is associated with a driver identifier | |
EP2836410B1 (en) | User identification and personalized vehicle settings management system | |
JP4267584B2 (en) | Device control apparatus and method | |
US20140266623A1 (en) | Systems, methods, and apparatus for learning the identity of an occupant of a vehicle | |
US20150046060A1 (en) | Method and System for Adjusting Vehicle Settings | |
CN110088761A (en) | Vehicle-mounted image authentication device | |
US11042614B2 (en) | Authentication device and authentication method | |
KR101165106B1 (en) | Driver certification device and method that do fetters in face realization | |
CN112638703B (en) | Seat adjusting method, device and system | |
CN114802293A (en) | Vehicle lateral control system with adjustable parameters | |
US20220375261A1 (en) | Driver recognition to control vehicle systems | |
JP2021146759A (en) | Movable body authentication system and movable body authentication program | |
CN111655547B (en) | Configuration adjustment method and device, movable platform and storage medium | |
CN111655547A (en) | Configuration adjustment method and device, movable platform and storage medium | |
CN116279238A (en) | Position adjustment method and device for driving assistance device and driving device | |
CN115071514A (en) | Vehicle and adjusting system for vehicle setting |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: E.SOLUTIONS GMBH, GERMANY Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:KOEHLER, THOMAS, DR.;STOCK, MATTHIAS;REEL/FRAME:051129/0334 Effective date: 20191121 |
|
FEPP | Fee payment procedure |
Free format text: ENTITY STATUS SET TO UNDISCOUNTED (ORIGINAL EVENT CODE: BIG.); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NOTICE OF ALLOWANCE MAILED -- APPLICATION RECEIVED IN OFFICE OF PUBLICATIONS |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
FEPP | Fee payment procedure |
Free format text: PETITION RELATED TO MAINTENANCE FEES GRANTED (ORIGINAL EVENT CODE: PTGR); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |