CN108161933A - Interactive mode selection method, system and reception robot - Google Patents
Interactive mode selection method, system and reception robot Download PDFInfo
- Publication number
- CN108161933A CN108161933A CN201711282942.3A CN201711282942A CN108161933A CN 108161933 A CN108161933 A CN 108161933A CN 201711282942 A CN201711282942 A CN 201711282942A CN 108161933 A CN108161933 A CN 108161933A
- Authority
- CN
- China
- Prior art keywords
- user
- age
- range
- information
- robot
- 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.)
- Pending
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J11/00—Manipulators not otherwise provided for
- B25J11/008—Manipulators for service tasks
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J11/00—Manipulators not otherwise provided for
- B25J11/0005—Manipulators having means for high-level communication with users, e.g. speech generator, face recognition means
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1602—Programme controls characterised by the control system, structure, architecture
- B25J9/161—Hardware, e.g. neural networks, fuzzy logic, interfaces, processor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/008—Artificial life, i.e. computing arrangements simulating life based on physical entities controlled by simulated intelligence so as to replicate intelligent life forms, e.g. based on robots replicating pets or humans in their appearance or behaviour
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Automation & Control Theory (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Biomedical Technology (AREA)
- Fuzzy Systems (AREA)
- Manipulator (AREA)
Abstract
Robot interactive mode selecting method is received the invention discloses a kind of, including：User is sensed to approach；Judge whether user belongs to registered user, languages and interactive mode are if it is selected according to user's registration information, if otherwise creating and recording user's registration information；Judge whether to receive user speech, languages are if it is selected according to user speech, if otherwise identification user classification；According to user's categorizing selection languages；Judge the age of user；Interactive mode is selected according to the age of user.Interactive mode selection method, system and reception robot using this method and/or system according to the present invention correct age of user on the basis of the facial characteristics of user using voice spectrum and/or physiological characteristic, different interactive modes is selected according to age of user, so as to improve the success rate of age of user prediction, user experience is improved.
Description
Technical field
The present invention relates to field in intelligent robotics, can judge age of user automatically more particularly to a kind of so as to select difference
Method, interactive mode selection system and the reception robot using this method and/or system of interactive mode.
Background technology
With the development of robot technology, robot is applied to every field, and existing robot is divided into two classes, i.e. work
Industry robot and specialized robot.So-called industrial robot is exactly towards the multi-joint manipulator of industrial circle or multiple degrees of freedom machine
Device people.And specialized robot is then in addition to industrial robot, for nonmanufacturing industry and serves the various advanced machines of the mankind
Device people, including：, underwater robot, amusement robot, military robot, agricultural robot, robotization machine etc..And it services
Robot is frequently utilized in the occasions welcomes such as bank, market, restaurant, the portion that sells house, hotel reception, guide service and advertising
Service trades are waited, this kind of service robot can set the operating modes such as welcome, inquiry, food delivery, checkout, amusement, server
Device people have the characteristics that intelligence substitute manpower, in addition, have the function of with people interaction, compared to the waiter in life, server
Device people can preferably try to please and customer is attracted either client etc. and to bring completely new service body to customer or client
It tests, while service robot can save the cost of labor of businessman；Can meet can also ensure that while working long hours it is excellent
Matter service avoids manual service and generates the tired feelings that the satisfaction of customer or client is caused to decline due to working long hours
Condition greatly improves work efficiency.
The application places of current service robot hotel for example concerning foreign affairs, airport duty-free shop, external mechanism, shopping mall
All it is using the service of pure Chinese operating system or reception robot Deng majority.Minority reception robot is using Chinese, English two
A parallel system, is then selected between the two systems.But due to technology barrier, dual system selection is restarted
The operations such as robot, complex steps take and extremely grow, in actual use, great inconvenience caused to user.And due to
Lack the support for other languages, the international personage for more grasping different language is made to have no way of selecting.
Further, robot uses built-in system voice（Usually young woman or male）It is interacted with user, example
User, the consulting for answering user are such as greeted, subscriber calendar arrangement is reminded, serves as venue explanation.However, existing system language
Single sound languages are for example only Chinese or English, and it is for example only young voice that pattern is single, can not be directed to different nationalities, no
User's progress specific aim interaction of same gender, all ages and classes.
A kind of possible improved procedure is to prejudge the year of user according to the voice sound ray characteristic frequency spectrum for collecting user
Age, however this single decision procedure is easily interfered.Such as when user's not sounding or more loud there are other at one's side
During source or as user because when sound mutation occurs for health reason or user emotion fluctuation or during the deliberately change of voice,
Judge that age of user becomes unreliable by user's sound spectrum.
Invention content
Therefore, it is an object of the invention to automatically judge age of user and different interaction moulds is selected according to age of user
Formula, so as to effectively improve user experience.
Robot interactive mode selecting method is received the present invention provides a kind of, including：User is sensed to approach；Judge to use
Whether family belongs to registered user, if it is languages and interactive mode is selected according to user's registration information, if otherwise created simultaneously
Record user's registration information；Judge whether to receive user speech, languages are if it is selected according to user speech, if otherwise
Identify user's classification；According to user's categorizing selection languages；Judge the age of user；Interactive mode is selected according to the age of user.
The step of wherein identification user classifies further comprises：According to biological information to user's rough sort；Belonged to according to auxiliary
Property to subscriber segmentation class, wherein, optionally, the biological information include Skin Color Information, face contour information, height, paces, shifting
Dynamic speed, limbs are static/walking posture；Optionally, secondary attribute includes the body on clothing style, user's carrying luggage or clothing
The language of part mark, the device name of the language in user-portable device operation interface, user-portable device；Optionally,
Rough sort includes yellow, white people, black race；Optionally, user's registration information includes the recognition of face information of user, user
Identity information, user language information, the age information of user.
Wherein, judge to further comprise before the age of user, receive user feedback and change user's registration information.
Wherein, the step of age for judging user, further comprises：The of age of user is predicted according to user's face feature
One range；Second range of age of user is generated according to user speech frequency spectrum and/or user year is generated according to user's physiological characteristic
The third range in age；Optionally, the 4th range is generated according to user's secondary attribute.
Wherein, if the second range is Chong Die with the first range or partly overlap, using the range of overlapping as age of user
Final range；If the second range is not be overlapped with the first range and third range is Chong Die with the first range or partly overlaps, by
Final range of the three ranges part Chong Die with the first range as age of user；If first, second, third range two is neither
Overlapping is not overlapped, and the part that the 4th range is Chong Die with any one of first, second, third range is as age of user
Final range.
Wherein, interactive mode includes respect language pattern, child mode, old pattern, default mode.
Wherein, respect language pattern is identical with volume, the word speed of default mode；Optionally, the volume of child mode and word speed are equal
Less than default mode；Optionally, the volume of old pattern is higher than default mode, and word speed is less than default mode.
Wherein, the volume of the ambient noise based on reception robot local environment in default mode, according to user and reception
The distance between robot, user age and volume, word speed are set.
The present invention also provides a kind of reception robot interactive mode switching system, for performing according to any of the above item institute
The method stated, including：User approaches close to sensing module for sensing user；User information registration module is used for recording
Family log-on message；User speech identification module, for receiving user speech；Age of user identification module, for judging user's
Age；Processor is used for：Judge whether user belongs to registered user, if it is according to user's registration information select languages and
Interactive mode, if otherwise being created in user information registration module and recording user's registration information；Judge whether to receive use
Family voice if it is selects languages according to user speech, if otherwise identification user classification, and according to user's categorizing selection language
Kind；And interactive mode is selected according to the age of user.
Invention additionally provides a kind of reception robots, are used using being received according to aforementioned any one of them method choice
Languages used in family and interactive mode.
According to the interactive mode selection method of the present invention, system and the reception robot using this method and/or system
Correct age of user using voice spectrum and/or physiological characteristic on the basis of the facial characteristics of user, according to age of user come
Different interactive modes is selected, so as to improve the success rate of age of user prediction, improves user experience.
Purpose of the present invention and other purposes unlisted herein, in the range of the application independent claims
It is satisfied.The embodiment of the present invention limits in the independent claim, and specific features limit in dependent claims thereto.
Description of the drawings
Carry out the technical solution that the present invention will be described in detail referring to the drawings, wherein：
Fig. 1 shows the schematic diagram of reception robot system according to embodiments of the present invention；
Fig. 2 shows the flow chart of reception robot interactive mode selecting method according to embodiments of the present invention；
Fig. 3 shows the flow for the specific steps that age of user is judged in flow chart shown in Fig. 2 according to embodiments of the present invention
Figure；And
Fig. 4 shows the block diagram of reception robot interactive model selection system according to embodiments of the present invention.
Specific embodiment
Carry out the feature and its skill of the present invention will be described in detail technical solution referring to the drawings and with reference to schematical embodiment
Art effect discloses the interactive mode selection method that can effectively improve user experience, system and uses this method and/or system
Reception robot.It should be pointed out that the structure that similar reference numeral expression is similar, term use herein " the
One ", " second ", " on ", " under " etc. can be used for modifying various system units or method and step.These modifications are unless stated otherwise
Space, order or the hierarchical relationship of institute's modification system component or method and step are not implied that.
As shown in Figure 1, reception robot according to embodiments of the present invention includes：Highly sensitive microphone 1 positioned at the crown,
For acquiring or receiving ambient enviroment acoustic information or personnel's voice messaging；Positioned at the high-definition camera 2 of forehead, for acquiring or
The topology information of reception staff's face（Such as bone contours）；Positioned at the fine sensor 3A and 3B of eye, for capturing
The face detail of personnel（Such as iris, retina, the dynamic change at eyebrow or canthus, the smile degree of lip or tooth reflection,
The slight twitch of ear or nose）So as to reflect the biological information of personnel or emotional information；It is touched positioned at robot various pieces
Sensor, including chin touch sensor 4, abdomen touch sensor 7, crown touch sensor 10, left ear/auris dextra touch sensing
Device 12A/12B, hindbrain touch sensor 13, left shoulder/right shoulder touch sensor 15A/15B, buttocks touch sensor 17, these are touched
Touch sensor for identify with the interaction of the sense of touch of user, so as to improve the accuracy for user identity or Emotion identification, and
It provides with the stress information of user's extremity to feed back, change movement/rotational parameters of robot body；Positioned at neck neck
3D depth cameras 5, for acquiring the depth of view information of surrounding scene；Positioned at the touch display screen 6 of chest, for show heartbeat or
The raw information of plan of colour of skin variation is so as to improve the fidelity of robot or broadening to traversing entire chest（It is not shown）And to
Family shows reception/Query Information or other video informations；Positioned at underbelly 2D laser radars 8, for measuring user or scene
In the distance of other mobile objects and robot, the height of auxiliary judgement object, movement speed, limbs are static/walking posture etc.；
Positioned at the Omni-mobile wheel 9 of foot, for the path movement for entire Robot being driven to prestore or real-time judgment selects；It is located at
The loud speaker 11A/11B of ear, for transmitting voice, audio-frequency information to user；Positioned at the emergency stop switch 14 of back part, for tight
Emergency stop stops the movement or action of robot, convenient for improving safety；Positioned at the power on button 16 of rear waist, connect for manually starting
The operating system of robot is treated to provide reception, counseling services；Positioned at the hand biosensor 18 of hand, for acquiring user
Fingerprint, measure user moisture content of skin（Resistivity）Or stress, the measurement user's pulse that roughness, measurement and user shake hands
Or oxygen content of capillary etc.；Charging interface 19 positioned at leg side and the power switch positioned at the leg back side.
Fig. 2 shows the flow chart of reception robot interactive mode selecting method according to embodiments of the present invention, and Fig. 4 is then shown
The block diagram of robot interactive model selection system is received used by having shown the embodiment.
First, it is approached by user close to sensing module to sense user.For example, pass through highly sensitive microphone 1, high definition
Camera 2,3D depth cameras 5 or 2D laser radars 8 or other approaching sensors（It is not shown, such as including bioelectricity
Field sensor, magnetic field sensor, sensor of chemical gas, mechanical vibration sensor）Receiving sensor information, if sensor is believed
Breath is more than preset threshold value（The storage device that selection system is included is obtained and is pre-stored according to high-volume test data（Do not show
Go out）In）Then judge in the neighbouring certain distance of robot or effective range（Such as 5 meters）In the presence of the personnel to the movement of reception robot
Or user.If there is close user, then reception robot system particularly speech selection system is waken up.If it is determined that it does not deposit
It is then receiving robot and is continuing to keep standby or dormant state, can so save electric energy, improving the continuation of the journey energy of reception robot
Power.
Then, judge whether user belongs to registered user by user information registration module, if it is according to user's registration
Information selection reception languages used and interactive mode, if otherwise creating and recording user's registration information and be stored in reception machine
The memory of device people（It is not shown）In.
User's registration information includes the recognition of face information of user, mainly includes facial skeleton profile（Topology）Structure, rainbow
Film/retinal feature information, facial surface characteristic information（Such as eyebrow distribution, eyelash length/curvature, the shape of color spot/mole
Shape/position etc.）, these recognition of face information create and recorded in the case where recognizing the user for the first time by reception robot.
User's registration information further includes the identity information of user, mainly comprising name, gender, height, native place/nationality, occupation etc., these
Subscriber identity information is actively provided during subsequent feedback by user or according to being consumed in venue, market, hotel etc.
It records and records automatically.User's registration information further comprises the language information of user, mother tongue, first foreign language comprising user,
Second Foreign Language, dialect etc., reception robot when recognizing user for the first time according to user be actively entered or system automatic identification
And it records, and can change during subsequent feedback or be corrected automatically using consumer record.User's registration information is into one
Step includes the age information of user, such as is manually entered including user or the accurate year of user using typings such as consumer records
What age or reception robot were predicted according to the facial characteristics of user, voice spectrum, physiological characteristic and other secondary attributes
The range of age of user.
User interaction patterns include：
1）Respect language pattern（There are the languages of respect language form for Japanese, Korean, Tibetan language etc. itself）, used by receiving robot
Reception voice usually selects young women or male, simulates the affiliated age bracket of the sound set as 20~25 years old.Therefore, if sentenced
The languages for determining user are the languages with respect language form, and the range of age of the accurate age of user or prediction is more than 20
Year, then using respect language form, volume, word speed are identical with default mode；If it is not, then the interactive mode using acquiescence；
2）Child mode, if it is decided that the age of user is less than reception robot application place the country one belongs to or area for children
Legal definition, such as 8 years old, 10 years old, 12 years old or 14 years old, in interactive language using children's pet phrase, for example add folded word or
Cartoon figure's onomatopoeia, and preferably the sound of young women as interactive voice, while reduces volume, slows down word speed to protect children
Youngster's hearing；
3）Old pattern, if it is decided that the age of user is higher than reception robot application place the country one belongs to or area for old age
Definition, such as 60 years old or 65 years old, improve volume, slow down word speed in order to which the old man of hearing loss can understand, obtain in time
Information；
4）Default mode, the volume of the ambient noise based on reception robot local environment, according to user and reception robot it
Between distance（The bigger volume of distance is bigger）, user age（Acquiescence 20~60 years old, the age is bigger, and word speed is slower）And it sets silent
The volume recognized（Such as higher than 10~20dB of ambient noise）, word speed（Such as 70~120% corresponding to news report word speed）.
Specifically, in a preferred embodiment of the invention, user is collected by high-definition camera 2, fine sensor 3A/3B
Facial information, be supplied to the processor of system（It is not shown）User information in recognition of face, with system storage is carried out to register
Module recognition of face information in pre-stored user's registration information make comparisons.
In other preferred embodiments of the present invention, by bio-electric field sensor, sensor of chemical gas or above-mentioned hand
Portion's biosensor 18 identifies the identity of user（Special bio-electric field either unique smell or hand it is relevant on
State other biological feature）Or by RFID tag reader（It is not shown）Identification carries the identity of the user of identity label.
If result of the comparison determines that user belongs to registered users, user's registration letter is extracted from system storage
Breath determines that reception robot is used for the user according to user recorded in user's registration information using language message
Corresponding languages, interactive mode are greeted, are seeked advice from, are received.
If result of the comparison determines that user is not belonging to registered users, the record of user's registration information is created, including
Recognition of face information, identity information, language information, age information.
Next, it is determined that user speech information whether is received, if（Such as user is during reception robot ambulation
It is talking with surrounding people or is being conversed with mobile phone with extraneous）Reception language is then selected according to the voice of user, if not（Example
Such as user in the webpage of browsing mobile phone or tablet computer, viewing video, listen melody with earphone and do not make a sound actively）Then
Further identification user classification.Specifically, it is acquired or received close to machine by the highly sensitive microphone 1 on the reception robot crown
People is certain or preset range in（Such as 2 meters, which, which is less than, judges user close to distance range used）User actively send out
The voice messaging gone out grasps language, such as Chinese, day by user speech identification module according to voice messaging discriminance analysis user
Text, Korean, English, French, Spanish, Arabic etc., and receive robot voice and select system that will use identification
To user grasp language interacted later with user, to user's active inquiry, reply consulting item etc..
Then, in the case that user does not make a sound actively, voice selecting system using robot acquired about
The non-audio information of user and with reference to big data statistical result to predict the classification of user, and select to use according to user's classification
The language that family may skillfully use.
For example, for the team such as travel party customer, it can make a reservation for move in the time of hotel or visit according to travel party, and
With reference to 3D depth cameras 5,2D laser radars 8（And the optional thermal imaging system not showed that）The team's number recognized
And judge that team corresponds to which of predetermined team's list, and belonging country or ground are retrieved according to the information of predetermined team
It is greeted so as to select the language of the country in area.
For example, for the tourist near airports shopping center, can be caught according to high-definition camera 2,3D depth cameras 5
The label tag that tourist's luggage case is wrapped is grasped, identify label tag plays enclave so as to judge tourist belonging country or area so as to pre-
Sentence the language that the tourist may skillfully use.
In another example the user for watching video, can be identified video caption or comment barrage by high-definition camera 2
Language（Chinese, English etc.）And select the subtitle or the corresponding language of barrage.
Wherein, in a preferred embodiment of the invention, the step of classifying for user further comprises, first according to life
Object information carries out rough sort to user, is then finely divided class to user according to secondary attribute.
In an embodiment of the invention, biological information includes the progress obtained colour of skin of recognition of face of high-definition camera 2
Information（Yellow, white, black）, face contour information（Face is wide/narrow, and nose is high/low, lower jaw side/point, Sunken orbital socket depth,
Cheekbone projecting height etc.）, height, paces measured by 2D laser radars 8, movement speed, limbs are static/walking posture etc..Root
Yellow will be roughly divided into according to these biological informations close to the user of reception robot（Anticipation may use Chinese, Japanese, Korea Spro
Language）, white people（Anticipation may use English, French, German, Spanish, Portuguese, Arabic）, black race（Anticipation
English, French, German, Spanish, Portuguese may be used）Three categories.
In a preferred embodiment of the invention, face/gesture recognition is carried out based on geometric properties.
First, the image data acquired to high-definition camera 2 pre-processes.Using the innovative variable neighborhood method of average
Smooth original image to eliminate most of noise, specifically, for w high h pixel unit, width pixel unit image g (x,
y)(0<=x<w,0<=y<H), the window of a n*n is taken out centered on being put using (i, j) as neighborhood, averaged rear output center pixel
Gray value g*(x,y).Usual n is simply selected to be a fixed odd number, such as 3,5 etc. in the prior art, for big
Majority application is enough to eliminate noise.However, for the scene that illumination condition is limited or atmospheric visibility is relatively low, image obscures journey
Degree can aggravate.Therefore, in a preferred embodiment of the invention, using smaller n1（Such as 3）Calculate initial pictures gray scale value set
{g*(x, y) }, then compare (i, j) and put corresponding gray value g*(x, y) is m with (i, j) point distance（Such as 1 ~ 4）Consecutive points
Initial gray value g*(x, y) ', the g if the two difference is within threshold value T*The final gray value that (x, y) is put as (i, j),
Larger n2 is used if difference is more than or equal to T（Such as 5 or 7, the odd number less than or equal to n1+m）Recalculate a small number of ashes
Degree fluctuates big specified point, so can obtain better smooth effect with smaller calculation amount, overcomes under haze, dark environment
Illumination, air cleanliness factor the problems such as.After preliminary smoothing processing, the minority being still had in image is isolated and is made an uproar
Sound, further using two dimension median filter denoising.Similar as before, the dimension n of two-dimentional window is equally variable, Ye Jixian
It is calculated for the first time using smaller n1, then compares the gray value of each point in distance m, if difference is in threshold value T2
It remains unchanged using first calculated value, window calculation gray scale is reset using larger n2 if difference is more than or equal to T2
Value.Change of scale is then carried out based on linear interpolation, and considers that average gray and variance carry out intensity profile markization, into one
Step is preferably based on Canny methods progress edge detection and binary conversion treatment finally obtains graphics set to be detected.
Then, recognition of face or gesture recognition are carried out based on detection geometric properties point（The variation of skeletal support frame）.At this
In invention preferred embodiment, the recognition of face facial characteristics based on geometric properties has the center of two, the exterior measuring side of nose
Edge position, corners of the mouth position, the supplementary features point constructed, particularly, the present invention also further uses eye width, open wiring and nose baseline
Spacing, auricle length and width are distinguished to refine the subspecies between same ethnic group（Such as South Korea's human eye width is small etc.）.Examine human face
Position when, it is first determined then two centre coordinates extract eyebrow according to the facial ratio of face on the basis of two positions
Each window such as window, eyes window, nose window, face window, auricle window carries out independent calculating, detects characteristic point namely calculates each
The position of organ and geometric parameter.It is similar, gesture recognition use head, shoulder, hip, ankle, elbow position and width with
And its change to characterize geometric properties point.In order to overcome the slight change of face/walking postures to the influence of recognition result, according to
Characteristic point constructs foundation of the feature vector with size, rotation and shift invariant as recognition of face.Specifically, (fix,
fiy) it is characterized point fiCoordinate, dij=√(fix-fjx)2+( fiy-fjy)2For distance between two points, then set of eigenvectors is combined into { d12/
d7a, d34/d7a, d23/d7a, d14/d7a, d56/d7a, d1a/d7a, d1a/d7a, d2a/d7a, d3a/d7a, d4a/d7a, d5a/d7a, d6a/d7a,
dba/d7a}.Then according to the feature vector of obtained all training samples training neural network classifier, such as Bagging god
Through network.
Further, using adaptive weighted multi-categorizer, the geometric properties point extraction subspace arrived for aforementioned identification is special
Sign obtains two single classifier taxonomic structures and exports last classification results using weighting coefficient estimation.And it further, is based on
Ethnic group classification results, with reference to reception robot historical data（Such as registered user's data, the use receiving and provided feedback
The data at family etc.）To prejudge the languages that user may select.
Secondary attribute includes the clothing style of user（Such as Chinese style Tang style clothing/cheongsam, Japanese kimonos, Arabic dust-robe, Su Ge
Blue skirt, ethnic group's characteristic clothes etc.）, user carry identity on luggage or clothing（Such as travel party's League flag, regimental
Or group's name on luggage）, user-portable device（Mobile phone, laptop, tablet computer etc.）Language in operation interface or
Corresponding language of bluetooth ID/ device names of person's user-portable device etc..Secondary attribute is equally by high-definition camera 2, essence
Thin sensor 3A/3B, 3D depth camera 5, which is combined, to be obtained.
Voice selecting system is received by the identification for user's secondary attribute, is carried out with the big data that network statistics obtains
It compares, prejudges the user for having identified secondary attribute in above-mentioned three categories user rough sort skillfully using each language
Probability is chosen user's most probable and is used（Probability highest）Language as reception robot will be interacted with the user used in language
Speech.Specifically, certain user is judged as white people and wears Arabic dust-robe then selecting Arabic, certain user is judged as Huang
It kind of people and wears kimonos and then selects Japanese, certain user is white people and wears Scotland skirt and then selects English, certain user is Black people
And cell phone apparatus number then selects French with French characters, certain user is white man and wears the young team uniform of rich card and then selects Spain
Language, certain user then select Sichuan dialect, etc. for yellow and travel party's name comprising " Sichuan ".
According to a preferred embodiment of the present invention, further comprise receiving user feedback after selection languages and change use
Family log-on message.Specifically, user feedback module receives user feedback, such as obtains user plane by fine sensor 3A/3B
The emotional information that portion's details is reflected（Such as surprised or smile）, pass through highly sensitive microphone 1 receive user direct audio
Feedback（Such as it praises or cries out in alarm）, pass through touch screen 6 obtain user touch feed back（Such as it chooses, draw and sweep）, pass through the crown
Touch sensor 10 or 18 grade touch sensors of hand biosensor obtain the force feedback of user（Such as it shakes hands, pat）, and
And these feedback informations are sent to user information registration module.When decision-feedback information is positive or positive, language will be received
The currently employed language of sound selection system is recorded in user's language information included in user's registration information.
More than, user's languages determined by user speech or non-speech data and select accordingly corresponding languages as
Language used by reception robot is interacted with user.Then, as described in Figure 3, the application is on the basis of the facial characteristics of user
Age of user is corrected using voice spectrum and/or physiological characteristic, different interactive modes is selected according to age of user.
First, the first range of age of user is predicted according to user's face feature.For example, using reception robot volume is located at
The high-definition camera 2 in portion, based on Flexible Model about Ecology（Flexible Models）, Snake active contour model（Active
Contour Model, ACM）, active shape model（Active Shaper Model, ASM), active apparent model (AAM) etc.
The profile information of method extraction user face.Here Image Multiscale analysis can be carried out using Two-Dimensional Gabor Wavelets, accurately
It extracts multiple images Block direction feature in image local area and carries out global feature extraction, Flexible Model about Ecology is recycled to carry out part special
Sign precisely extraction.Then, features of skin colors extraction is carried out for the image or video of the acquisition of high-definition camera 2, is obtained from video
One frame image is detected, can with timing acquisition image or have been detected by scene have personage when timing acquisition list note image
It is handled；Image preprocessing because skin cluster is different with face extraction, needs to carry out Preprocessing, will not
Remove in the region for belonging to skin；Features of skin colors obtains, and by skin color segmentation model, will know closest to the features of skin colors of human body
Not.Its method has simple color and position clustering algorithm, color and position to cluster preferential Graph-Cut algorithms etc., color
It is preferable that preferential Graph-Cut algorithm effects are clustered with position.Then, it using the fine sensor 3A and 3B positioned at eye, obtains
The minutia of user's face is taken, such as canthus, brow furrows are distributed and number, facial skin smoothness/reflectance, the corners of the mouth,
Ptosis degree, the neighbouring cutis laxa degree of cheekbone etc..
On the basis of user's face mask achieved above, skin color, face detail, the big data with network acquisition
Or the database to prestore is compared, and prejudges the first range of age of user.For example, user is the yellow race and skin
Relaxation degree/wrinkle distribution density is more than 50% of average value and judges that age bracket is 60 years old or more, user for the white race and
Cutis laxa degree/wrinkle distribution density then judges that age bracket is 60 years old or more more than average value 30%, and user is the white race
And cutis laxa degree/wrinkle distribution density subaverage 40% then judges age bracket for under-18s, user is black people
Kind and cutis laxa degree/judgement age bracket under-18s of wrinkle distribution density subaverage 30% etc..
Then, the first range is corrected according to user speech frequency spectrum and generates the second range.If judgement reception machine shown in Fig. 2
Result the step of whether device people receives voice obtains be "Yes", according to user speech select languages while or it
Afterwards, the processor of robot is received for personnel's voice for acquiring or receive via the highly sensitive microphone 1 positioned at the crown
Information is analyzed to obtain frequency spectrum.Specifically, Short Time Fourier Transform is made to voice signal, obtains its power spectrum chart；From
The frequency corresponding to some representative wave crests is found out in its power spectrum chart, a spy is made with these characteristic frequency values
Sign vector；With some particular person sound namely reference voice（Such as the young women to prestore or the man of reception robot interactive
The sound of property or the sound of newscaster）Feature vector for standard vector F, define the feature vector G of sound to be measured with
The distance between F function D, the corresponding D of different user to prestore according to database and the correlation between its age set multiple
（At least two）Threshold value P1、P2、P3……PN.When D is less than or equal to P1When, judgement user is children, is, for example, less than 12 years old；When D is big
In P1And less than or equal to P2When, judgement user is teenager, such as 13 to 16 years old；When D is more than P2And less than or equal to P3When, judgement is used
Family is youth, is more than P as D within such as 17 to 25 years old ...N-1And less than or equal to PNWhen, judgement user is old age, such as 60~80
Year；When D is more than PNWhen, judgement user is man at an advanced age, such as more than 80 years old, etc..
In the process, if according to audio judge the second range fall into according to facial characteristics judge first in the range of
Or partly overlap, then using the range identification of overlapping as correct the range of age.If the second range and the first range are complete
It is misaligned, then the two ranges are pre-stored in register（It is not shown）In to treat subsequently to continue with.
If it is shown in Fig. 2 judge to receive result that robot the step of whether receiving voice obtains for "No" namely
Reception robot, which does not recognize, nearby has user to speak（Including user's not sounding or distant, sound is relatively low and nothing
Method accurately identifies）, then processor according to user's physiological characteristic correct age of user the first range.
Specifically, reception robot actively attracts user to touch the hand biosensor 18 positioned at hand, for acquiring
The fingerprint of user, the moisture content of skin for measuring user（Resistivity）Or roughness, measurement and user shake hands stress, measure user
Oxygen content of pulse or capillary etc..The mode of attraction can be voice prompt, display screen Subtitle Demonstration, interactive dancing etc.
Deng.Biosensor 18 is preferably skin-texture detection device, corresponding to judge for detecting the dermatoglyph of user's hand
Age range.The mode for detecting dermatoglyph can be by silicon wafer（Solid-state）Identifier is reached, and is to utilize small capacitor
To sense the texture fold of skin；Or optical scanner, such as use many charge coupled cells（CCD）The array of composition comes
The digital image of skin is acquired, light emitting diode can also be used（LED）The image of light source irradiation skin back reflection carrys out analyzing skin
Texture.Current research finds dermatoglyph, such as degree of roughness, wrinkle depth, skin mound（Or skin ridge, refer to the protrusion of skin surface
Position）With sulci of skin（The depressed area of skin surface）Between distance, can substantially reflect different age levels.It therefore, can basis
Cutaneous roughness, rill depth, i.e. physiological characteristic judge the age range of user.
For example, less than user is judged if first threshold for children be, for example, less than 12 years old if the skin roughness of user, skin
Skin roughness but is below second threshold higher than first threshold and is then determined as teenager such as 13~25 years old, and skin roughness is higher than
Second threshold and then it is determined as such as 26~40 years old middle age less than third threshold value, skin roughness is higher than third threshold value and less than the
Four threshold values are then determined as person in middle and old age such as 41~60 years old, and skin roughness is then determined as old age such as 61~80 higher than the 4th threshold value
Year etc..
【In another preferred embodiment of the present invention, the micro- biography of integrated protein is further included in hand biosensor 18
Sensor, for detecting on skin the protein with tissue, cell or subcellular specificity to reflect the age of user indirectly.Example
Such as, microsensor includes the detecting electrode by being formed in oxide semiconductor substrate with respect to two graphenes, works as specific proteins
Due to electric polarity and rotational symmetry that protein macromolecule is different when matter is across detecting electrode, subthreshold value electricity between electrode
Stream can occur to change accordingly so as to corresponding to different proteins type.The microsensor is exclusively used in detection aquaporin matter,
Belong to water-glycerol channel protein subtribe, it is neutral also for glycerine and urea etc. other than there is high-permeability for hydrone
Small molecule has high-permeability.Aquaporin participates in a variety of physiology of human body, pathologic process, concentration, air flue table with urine
The fluid balance in face has substantial connection.And the missing of this aquaporin can lead to dry skin, cuticula aquation
Decline, skin elasticity decline etc. reflect the process of skin aging.Particularly, it is examined using the micro protein matter sensor of the application
The expression quantity of aquaporin matter and mRNA expressions in user's skin histology are surveyed, difference is selected according to high-volume test set
Threshold value selects the threshold value Tm1 to be for 0.31, Tm2 0.55, Tm3 to reflect different age group, such as mRNA expressions
0.67th, Tm4 0.74, for aquaporin matter expression quantity select threshold value Ta1 for 2.03, Ta2 1.95, Ta3 1.88,
Ta4 is 1.21, children's group（Less than 12 years old）For Tm2 ~ Tm3 and less than Ta1, teenager's group（13 ~ 25 years old）For Tm3 ~ Tm4 and Ta1 ~
Ta2, middle aged group（26 ~ 40 years old）To be more than Tm4 and Ta2 ~ Ta3, person in middle and old age's group（41 ~ 60 years old）For Tm2 ~ Tm1 and Ta3 ~ Ta4, always
Year group（60 years old or more）To be less than Tm1 and more than Ta4.
In the another preferred embodiment of the present invention, the age of user is detected using the telomeric dna of skin.With in identification
It is similar to state specific proteins, using graphene counter-electrodes in integrated microsensor come analyze DNA macromoleculars across electrode it
Between caused curent change so as to react different attributes, such as DNA fragmentation length（Detect current pulse width or half height
It is wide）.Generally, DNA fragmentation length is gradually shortened with the increase at age, and very fast in the rate that adolescence shortens
And the short rate of retraction of growing up is slower.In the preferred embodiment of the application, the age of user Y that is obtained according to pre-stored data collection
（Year）With skin telomere DNA fragmentation length value X（Kb）Between relationship be Y=7.89/X2+ 55.419/X+5.047, therefore can be with
The X values of reaction are measured according to the microsensor of reception robot hand to be derived from the actual age of user in real time.
In other preferred embodiments of the invention, characterize user's using other skin parameters such as sebum, moisture, pH value
Age.The measurement of sebum is based on photometer principle, and moisture uses capacitance method, and pH value measurement is passed through based on H ions
Voltage difference caused by semi-permeable membrane.With the increase at age（In the range of identical BMI values）, the buffer capacity of skin is gradually reduced,
Sebum, moisture are gradually reduced, and pH value is gradually risen to close neutrality, therefore similar can be measured according to control group
Correlation between parameters and age sets different threshold values so as to judge the age distribution of active user.
When the range of age judged according to user's physiological characteristic（Third range）Or part weight Chong Die with aforementioned first range
When folded, using the range of overlapping as correct the range of age.If be not overlapped, by the first range with being sentenced according to user's physiological characteristic
Fixed the range of age record remains subsequent processing in a register.
In a preferred embodiment of the invention, optionally, it receives user speech and the second range is generated according to voice spectrum
Afterwards, third range further is generated using above-mentioned steps namely according to user's physiological characteristic, for correcting the second range.
Finally, the 4th range is generated according to user's secondary attribute to correct first/second/third range.It is as previously mentioned, auxiliary
Help attribute equally by high-definition camera 2, fine sensor 3A/3B, 3D depth camera 5 and RFID card reader（It is not shown）
Etc. joints obtain.Secondary attribute includes the clothing style of user（Such as Chinese style Tang style clothing/cheongsam, Japanese kimonos, Arabic dust-robe,
Scotland skirt, ethnic group's characteristic clothes etc.）, user carry identity on luggage or clothing（Such as travel party's League flag,
Group's name on regimental or luggage）, user-portable device（Mobile phone, laptop, tablet computer etc.）Language in operation interface
Corresponding language of bluetooth ID/ device names of speech or user-portable device etc..In addition, for correcting third range
User's secondary attribute still further comprises, and can reflect accessories, the dressing of age of user（Such as earrings/ear nail/tie clip style,
Lip gloss popular colour, the cartoon character carried are made comparisons with the age of user distribution of data-base recording）, can reflect user's receipts
Enter/the article tag of the level of consumption（Such as the brand of western-style clothes, wind coat, shoes and hats, age of user-income phase with data-base recording
Sex-intergrade is closed to compare）Etc..If the age of user judged according to user's secondary attribute（4th range）Third range is fallen into, then
Using the age of user as final the range of age.If first/second/third range two is neither overlapped, use according to
The Chong Die part of any one of the age of user of family secondary attribute judgement and first/second/third range（It is preferred that overlapping number is most
Part, it is such as be overlapped or be overlapped with first, third with first, second）As the final the range of age of user.
In this way, above scheme is received the video of the sensor acquisition of robot by comprehensive utilization, audio, touched
Feel, RFID information carry out the range of age of comprehensive descision user, improve the success rate of skeleton growth rings, and according to the age of user
Select different interactive modes, especially for Japan and Korea S customer using dedicated respect language form, for children/old man using adjusting
Volume/word speed of system etc., improves user experience, is conducive to improve the consulting performance of reception robot.
According to the interactive mode selection method of the present invention, system and the reception robot using this method and/or system
Correct age of user using voice spectrum and/or physiological characteristic on the basis of the facial characteristics of user, according to age of user come
Different interactive modes is selected, so as to improve the success rate of age of user prediction, improves user experience.
Although illustrating the present invention with reference to one or more exemplary embodiments, those skilled in the art could be aware that need not
It is detached from the scope of the invention and various suitable changes and equivalents is made to system, method.It in addition, can by disclosed introduction
The modification of particular condition or material can be can be adapted to without departing from the scope of the invention by making many.Therefore, the purpose of the present invention does not exist
In being limited to as the preferred forms for being used to implement the present invention and disclosed specific embodiment, and disclosed system and side
Method is by all embodiments including falling within the scope of the present invention.
Claims (10)
1. a kind of reception robot interactive mode selecting method, including：
User is sensed to approach；
Judge whether user belongs to registered user, languages and interactive mode are if it is selected according to user's registration information, if
Otherwise it creates and records user's registration information；
Judge whether to receive user speech, languages are if it is selected according to user speech, if otherwise identification user classification；
According to user's categorizing selection languages；
Judge the age of user；
Interactive mode is selected according to the age of user.
2. according to the method described in claim 1, the step of wherein identification user classifies further comprises：
According to biological information to user's rough sort；
According to secondary attribute to subscriber segmentation class,
Wherein, optionally, the biological information includes Skin Color Information, face contour information, height, paces, movement speed, limbs
Static/walking posture；Optionally, secondary attribute includes clothing style, user carries luggage or the identity on clothing, user
The language of the device name of language, user-portable device in mancarried device operation interface；Optionally, rough sort includes Huang
Kind people, white people, black race；
Optionally, user's registration information include the recognition of face information of user, the identity information of user, user language information,
The age information of user.
3. according to the method described in claim 1, wherein, judge to further comprise before the age of user, receive user feedback
And change user's registration information.
4. according to the method described in claim 1, wherein, the step of age for judging user, further comprises：
The first range of age of user is predicted according to user's face feature；
Second range of age of user is generated according to user speech frequency spectrum and/or age of user is generated according to user's physiological characteristic
Third range；
Optionally, the 4th range is generated according to user's secondary attribute.
5. it according to the method described in claim 4, wherein, if the second range is Chong Die with the first range or partly overlaps, will weigh
Final range of the folded range as age of user；If the second range is not be overlapped with the first range and third range and first
Range is overlapped or partly overlaps, and the part that third range is Chong Die with the first range is as the final range of age of user；If
First, second, third range two is neither overlapped or is not overlapped, by any of the 4th range and first, second, third range
Final range of the part of item overlapping as age of user.
6. according to the method described in claim 1, wherein, interactive mode includes respect language pattern, child mode, old pattern, writes from memory
Recognize pattern.
7. according to the method described in claim 6, wherein, respect language pattern is identical with volume, the word speed of default mode；Optionally,
The volume and word speed of child mode are below default mode；Optionally, the volume of old pattern is higher than default mode, and word speed is less than
Default mode.
8. according to the method described in claim 6, wherein, the ambient noise based on reception robot local environment in default mode
Volume, according to user and reception the distance between robot, user age and volume, word speed are set.
9. a kind of reception robot interactive mode switching system, for performing according to claim 1 to 8 any one of them side
Method, including：
User approaches close to sensing module for sensing user；
User information registration module, for recording user's registration information；
User speech identification module, for receiving user speech；
Age of user identification module, for judging the age of user；
Processor is used for：
Judge whether user belongs to registered user, languages and interactive mode are if it is selected according to user's registration information, if
Otherwise it is created in user information registration module and records user's registration information；
Judge whether to receive user speech, languages are if it is selected according to user speech, if otherwise identification user classification,
And according to user's categorizing selection languages；And
Interactive mode is selected according to the age of user.
10. a kind of reception robot is used using user is received according to claim 1 to 8 any one of them method choice
Languages and interactive mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711282942.3A CN108161933A (en) | 2017-12-07 | 2017-12-07 | Interactive mode selection method, system and reception robot |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711282942.3A CN108161933A (en) | 2017-12-07 | 2017-12-07 | Interactive mode selection method, system and reception robot |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108161933A true CN108161933A (en) | 2018-06-15 |
Family
ID=62524455
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711282942.3A Pending CN108161933A (en) | 2017-12-07 | 2017-12-07 | Interactive mode selection method, system and reception robot |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108161933A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109093631A (en) * | 2018-09-10 | 2018-12-28 | 中国科学技术大学 | A kind of service robot awakening method and device |
| CN109741744A (en) * | 2019-01-14 | 2019-05-10 | 博拉网络股份有限公司 | AI robot dialog control method and system based on big data search |
| CN109949795A (en) * | 2019-03-18 | 2019-06-28 | 北京猎户星空科技有限公司 | A kind of method and device of control smart machine interaction |
| CN110148399A (en) * | 2019-05-06 | 2019-08-20 | 北京猎户星空科技有限公司 | A kind of control method of smart machine, device, equipment and medium |
| CN110209957A (en) * | 2019-06-06 | 2019-09-06 | 北京猎户星空科技有限公司 | Explanation method, apparatus, equipment and storage medium based on intelligent robot |
| CN110427462A (en) * | 2019-08-06 | 2019-11-08 | 北京云迹科技有限公司 | With method, apparatus, storage medium and the service robot of user interaction |
| CN110610703A (en) * | 2019-07-26 | 2019-12-24 | 深圳壹账通智能科技有限公司 | Speech output method, device, robot and medium based on robot recognition |
| CN110913074A (en) * | 2019-11-28 | 2020-03-24 | 北京小米移动软件有限公司 | Sight distance adjusting method and device, mobile equipment and storage medium |
| CN111089581A (en) * | 2018-10-24 | 2020-05-01 | 上海博泰悦臻网络技术服务有限公司 | Traffic guidance method, terminal and robot |
| CN111506377A (en) * | 2020-04-16 | 2020-08-07 | 上海茂声智能科技有限公司 | Language switching method and device and voice service terminal |
| CN112929502A (en) * | 2021-02-05 | 2021-06-08 | 国家电网有限公司客户服务中心 | Voice recognition method and system based on electric power customer service |
| CN114633267A (en) * | 2022-03-17 | 2022-06-17 | 上海擎朗智能科技有限公司 | Interactive content determination method, mobile equipment, device and storage medium |
| CN115171284A (en) * | 2022-07-01 | 2022-10-11 | 国网汇通金财（北京）信息科技有限公司 | Old people care method and device |
| CN116787469A (en) * | 2023-08-29 | 2023-09-22 | 四川数智宗医机器人有限公司 | Digital traditional Chinese medicine robot pre-consultation system based on artificial intelligence |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004198831A (en) * | 2002-12-19 | 2004-07-15 | Sony Corp | Method, program, and recording medium for speech recognition |
| CN101618542A (en) * | 2009-07-24 | 2010-01-06 | 塔米智能科技（北京）有限公司 | System and method for welcoming guest by intelligent robot |
| CN103533438A (en) * | 2013-03-19 | 2014-01-22 | Tcl集团股份有限公司 | Clothing push method and system based on intelligent television |
| CN204723761U (en) * | 2015-01-04 | 2015-10-28 | 玉林师范学院 | Based on libraries of the universities' guest-meeting robot of RFID |
| CN106504743A (en) * | 2016-11-14 | 2017-03-15 | 北京光年无限科技有限公司 | A kind of interactive voice output intent and robot for intelligent robot |
| CN106649290A (en) * | 2016-12-21 | 2017-05-10 | 上海木爷机器人技术有限公司 | Speech translation method and system |
| CN106952648A (en) * | 2017-02-17 | 2017-07-14 | 北京光年无限科技有限公司 | A kind of output intent and robot for robot |
| CN107316254A (en) * | 2017-08-01 | 2017-11-03 | 深圳市益廷科技有限公司 | A kind of hotel service method |
| CN107358949A (en) * | 2017-05-27 | 2017-11-17 | 芜湖星途机器人科技有限公司 | Robot sounding automatic adjustment system |
-
2017
- 2017-12-07 CN CN201711282942.3A patent/CN108161933A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004198831A (en) * | 2002-12-19 | 2004-07-15 | Sony Corp | Method, program, and recording medium for speech recognition |
| CN101618542A (en) * | 2009-07-24 | 2010-01-06 | 塔米智能科技（北京）有限公司 | System and method for welcoming guest by intelligent robot |
| CN103533438A (en) * | 2013-03-19 | 2014-01-22 | Tcl集团股份有限公司 | Clothing push method and system based on intelligent television |
| CN204723761U (en) * | 2015-01-04 | 2015-10-28 | 玉林师范学院 | Based on libraries of the universities' guest-meeting robot of RFID |
| CN106504743A (en) * | 2016-11-14 | 2017-03-15 | 北京光年无限科技有限公司 | A kind of interactive voice output intent and robot for intelligent robot |
| CN106649290A (en) * | 2016-12-21 | 2017-05-10 | 上海木爷机器人技术有限公司 | Speech translation method and system |
| CN106952648A (en) * | 2017-02-17 | 2017-07-14 | 北京光年无限科技有限公司 | A kind of output intent and robot for robot |
| CN107358949A (en) * | 2017-05-27 | 2017-11-17 | 芜湖星途机器人科技有限公司 | Robot sounding automatic adjustment system |
| CN107316254A (en) * | 2017-08-01 | 2017-11-03 | 深圳市益廷科技有限公司 | A kind of hotel service method |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109093631A (en) * | 2018-09-10 | 2018-12-28 | 中国科学技术大学 | A kind of service robot awakening method and device |
| CN111089581A (en) * | 2018-10-24 | 2020-05-01 | 上海博泰悦臻网络技术服务有限公司 | Traffic guidance method, terminal and robot |
| CN109741744A (en) * | 2019-01-14 | 2019-05-10 | 博拉网络股份有限公司 | AI robot dialog control method and system based on big data search |
| CN109741744B (en) * | 2019-01-14 | 2021-03-09 | 博拉网络股份有限公司 | AI robot conversation control method and system based on big data search |
| CN109949795A (en) * | 2019-03-18 | 2019-06-28 | 北京猎户星空科技有限公司 | A kind of method and device of control smart machine interaction |
| CN110148399A (en) * | 2019-05-06 | 2019-08-20 | 北京猎户星空科技有限公司 | A kind of control method of smart machine, device, equipment and medium |
| CN110209957A (en) * | 2019-06-06 | 2019-09-06 | 北京猎户星空科技有限公司 | Explanation method, apparatus, equipment and storage medium based on intelligent robot |
| CN110610703A (en) * | 2019-07-26 | 2019-12-24 | 深圳壹账通智能科技有限公司 | Speech output method, device, robot and medium based on robot recognition |
| CN110427462A (en) * | 2019-08-06 | 2019-11-08 | 北京云迹科技有限公司 | With method, apparatus, storage medium and the service robot of user interaction |
| CN110913074A (en) * | 2019-11-28 | 2020-03-24 | 北京小米移动软件有限公司 | Sight distance adjusting method and device, mobile equipment and storage medium |
| CN111506377A (en) * | 2020-04-16 | 2020-08-07 | 上海茂声智能科技有限公司 | Language switching method and device and voice service terminal |
| CN112929502A (en) * | 2021-02-05 | 2021-06-08 | 国家电网有限公司客户服务中心 | Voice recognition method and system based on electric power customer service |
| CN114633267A (en) * | 2022-03-17 | 2022-06-17 | 上海擎朗智能科技有限公司 | Interactive content determination method, mobile equipment, device and storage medium |
| CN115171284A (en) * | 2022-07-01 | 2022-10-11 | 国网汇通金财（北京）信息科技有限公司 | Old people care method and device |
| CN115171284B (en) * | 2022-07-01 | 2023-12-26 | 国网汇通金财（北京）信息科技有限公司 | Senior caring method and device |
| CN116787469A (en) * | 2023-08-29 | 2023-09-22 | 四川数智宗医机器人有限公司 | Digital traditional Chinese medicine robot pre-consultation system based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108161933A (en) | Interactive mode selection method, system and reception robot | |
| CN108153169A (en) | Guide to visitors mode switching method, system and guide to visitors robot | |
| CN108182098A (en) | Receive speech selection method, system and reception robot | |
| US11989340B2 (en) | Systems, methods, apparatuses and devices for detecting facial expression and for tracking movement and location in at least one of a virtual and augmented reality system | |
| US11748980B2 (en) | Makeup evaluation system and operating method thereof | |
| JP6850723B2 (en) | Facial expression identification system, facial expression identification method and facial expression identification program | |
| CN112784763B (en) | Expression recognition method and system based on local and overall feature adaptive fusion | |
| US10062163B2 (en) | Health information service system | |
| KR20160012902A (en) | Method and device for playing advertisements based on associated information between audiences | |
| Liu et al. | Region based parallel hierarchy convolutional neural network for automatic facial nerve paralysis evaluation | |
| CN101305913B (en) | Face beauty assessment method based on video | |
| US20170352351A1 (en) | Communication robot | |
| EP3579176A1 (en) | Makeup evaluation system and operation method thereof | |
| US10423978B2 (en) | Method and device for playing advertisements based on relationship information between viewers | |
| Abobakr et al. | Rgb-d fall detection via deep residual convolutional lstm networks | |
| KR102316723B1 (en) | Body-tailored coordinator system using artificial intelligence | |
| CN109670406B (en) | Non-contact emotion recognition method for game user by combining heart rate and facial expression | |
| CN108198159A (en) | A kind of image processing method, mobile terminal and computer readable storage medium | |
| CN107316333A (en) | It is a kind of to automatically generate the method for day overflowing portrait | |
| CN113920568B (en) | Face and human body posture emotion recognition method based on video image | |
| US20090033622A1 (en) | Smartscope/smartshelf | |
| CN106599785A (en) | Method and device for building human body 3D feature identity information database | |
| CN113343860A (en) | Bimodal fusion emotion recognition method based on video image and voice | |
| CN112101235B (en) | Old people behavior identification and detection method based on old people behavior characteristics | |
| CN111967324A (en) | Dressing system with intelligent identification function and identification method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180615 |