CN108810838A - The room-level localization method known based on smart mobile phone room background phonoreception - Google Patents
The room-level localization method known based on smart mobile phone room background phonoreception Download PDFInfo
- Publication number
- CN108810838A CN108810838A CN201810560130.9A CN201810560130A CN108810838A CN 108810838 A CN108810838 A CN 108810838A CN 201810560130 A CN201810560130 A CN 201810560130A CN 108810838 A CN108810838 A CN 108810838A
- Authority
- CN
- China
- Prior art keywords
- room
- background sound
- background
- mobile phone
- smart mobile
- 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
- 238000000034 method Methods 0.000 title claims abstract description 33
- 230000004807 localization Effects 0.000 title claims abstract description 15
- 238000000605 extraction Methods 0.000 claims description 13
- 230000000644 propagated effect Effects 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000001228 spectrum Methods 0.000 claims description 7
- 238000009432 framing Methods 0.000 claims description 6
- 210000002569 neuron Anatomy 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 230000005236 sound signal Effects 0.000 claims description 6
- 238000013135 deep learning Methods 0.000 claims description 4
- 230000021615 conjugation Effects 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 230000001256 tonic effect Effects 0.000 claims 1
- 230000008676 import Effects 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 description 8
- 210000004027 cell Anatomy 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000008447 perception Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000002045 lasting effect Effects 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000035939 shock Effects 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 238000002604 ultrasonography Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/33—Services specially adapted for particular environments, situations or purposes for indoor environments, e.g. buildings
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
- G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
- G01S5/0252—Radio frequency fingerprinting
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
- G10L25/21—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters the extracted parameters being power information
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/27—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the analysis technique
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/45—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of analysis window
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Human Computer Interaction (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Mobile Radio Communication Systems (AREA)
- Telephone Function (AREA)
Abstract
The invention discloses a kind of room-level localization methods known based on smart mobile phone room background phonoreception, including position two stages on acquisition and training and background sound ray in place to be positioned under indoor room background sound ray.Indoor environment sound is enrolled using smart mobile phone, the 5th percentile power by extracting sound power is used as acoustic feature, import RNN-LSTM learning algorithms, training obtains background sound location model in given chamber, the discrimination in room can be calculated by being compared by the room information with true environment, identify that room-level positioning can be realized in room.For relatively traditional similar acoustic signature indoor locating system, the method for the present invention does not only reach the requirement of room-level positioning, and improves room discrimination, is more suitable in room background acoustic fix ranging scene.
Description
Technical field
It is specifically a kind of to be known based on smart mobile phone room background phonoreception the present invention relates to the localization method of indoor room grade
Room-level localization method.
Background technology
GPS be representative location technology since appearance, characteristic efficient with its, rapid, accurate makes people's lives
Drastic change occurs for mode, while its service having been driven to flourish with what is applied, brings great convenience to people's lives.But
It is traditional outdoor positioning technology (such as GPS) due to the limitation of principle, performance in room conditions is ideal not to the utmost, therefore
It is badly in need of efficient, convenient, the accurate indoor positioning technologies of one kind to plug a gap.
It is currently more ripe to have based on indoor positioning technologies such as Wifi, bluetooth, infrared ray, ultrasonic waves.Determined based on WiFi
Position technical foundation equipment is easily installed, but easily higher by other signals interference, power consumption；Location technology based on bluetooth is low in energy consumption,
Easy of integration, but orientation distance is short, stability is poor, easily by noise jamming；It is high based on infrared location technology precision, but cannot wear
Obstacle-overpass, while cost is high, power consumption is larger；Indoor positioning technologies overall precision based on ultrasound is high, simple in structure, but exists
Multipath effect, decaying is apparent, is easily affected by temperature, is of high cost.
Technological merit based on background acoustic fix ranging is easily acquired without the other infrastructure of pre-arranged and background sound.
In fact, background sound is distributed special Acoustic Wave Propagation form as a kind of time and space, human auditory system is acted on, it can shape
At the Auditory Perception effect of certain rule.Meanwhile background sound is also a kind of information carrier, reflects the physical attribute, outer of sounding body
The key property of numerous environmental factors such as portion's exciting force.In addition, architectural acoustics field proposes:The lasting sound in room and room
Shock response is combined the unique background sound for foring each room.Even two rooms similar in human auditory system, due to
Persistence sound caused by room unit remains able to more accurately distinguish two different rooms.Therefore using background sound into
Row positioning is feasible.
Oneself has part interior fingerprint location system, and advantage, acquisition WiFi, sound, vision figure are sensed using smart mobile phone more
Picture, accelerometer data carry out the fusion positioning of multi information as fingerprint；A small number of documents then specialize in indoor environment background sound
The method of positioning, such as：Background sound indoor positioning etc. is carried out by background sound fingerprint extraction and KNN algorithms.However, by different acoustics
Feature and the influence for not identifying sorting algorithm in unison, positioning accuracy are generally relatively low.
Invention content
The shortcomings that needing pre- deployment base facility for traditional indoor positioning, the present invention provide a kind of based on smart mobile phone room
The room-level localization method of interior background sound perception, it is only necessary to acquire indoor room background sound using smart mobile phone, extract background sound
Fingerprint simultaneously establishes background acoustic model；One is trained by RNN-LSTM learning algorithms to be suitable for determining under room background sound field scape
Bit model is applied to the positioning of indoor room grade.
The present invention is based on the room-level localization methods that smart mobile phone room background phonoreception is known, including (1) indoor room background
It acquires and trains under sound ray, and position two stages on (2) background sound ray in place to be positioned.
Acquisition and training, specific method include the following steps under stage (1) the indoor room background sound ray:
(1.1) indoor room background sound and feature extraction are acquired：
Using enough room background sound data are acquired under smart mobile phone line, background acoustic feature extraction is carried out, passes through the 5th
Percentile power draw goes out background sound fingerprint；
(1.2) background sound fingerprint base is built：
Background sound fingerprint is collectively formed into room background sound fingerprint base with the room label information manually marked；
(1.3) training process：
After constructing background sound fingerprint base, as training set data, trained by RNN-LSTM deep learning algorithms
Go out to be suitable for the location model under background sound indoor positioning scene, this model there need to be higher generalization ability, can preferably reflect
The feature of entire sample space.
It is positioned on stage (2) the background sound ray in place to be positioned, specific method includes the following steps:
(2.1) background sound test set finger print data in place to be positioned is obtained：
Place background sound to be positioned in studio carries out the 5th percentage power draw, and the background sound fingerprint obtained is as survey
Examination collection data；
(2.2) it by the background sound location model of training under test set Data In-Line, after input, exports to mark for room and believe
Breath, the discrimination that can calculate room is compared by the room information with true environment, identifies that room-level can be realized in room
Positioning.
The present invention establishes room background sound by the 5th percentile power draw background sound fingerprint using RNN-LSTM algorithms
Location model so that room discrimination has promotion by a relatively large margin.
Step (1.1) the 5th percentile power draw goes out background sound fingerprint, includes the following steps：
(1.1.1) carries out framing windowing operation to the original audio sequence of acquisition, obtains the background acoustical signal of short-term stationarity,
Window function is：
Each frame audio signal after framing adding window is done FFT transform by (1.1.2), retain FFT transform preceding two/
One data, and it is multiplied by its conjugation, power spectrum can be found out；
FFT transform formula is：
(1.1.3) gives up the audio signal that frequency is more than 7kHz；
(1.1.4) is ranked up remaining data by watt level；
5th percentage of (1.1.5) extraction power arranges and takes logarithm, obtains background sound fingerprint.
The 5th percentage power draw of carry out described in step (2.1), method are identical as step (1.1).First two steps are marks
Accurate spectra calculation method.After finding out power spectrum, need the feature vector that robustness is high in extraction power spectrum to characterize room
Between background sound.Since want extraction is background sound in room, this feature should have time stationarity, it is therefore desirable to inhibit
Transient noise.During window sample background is extracted by selecting the minimum value for the background acoustical power observed under each frequency
Sound spectrum.However, minimum value is easy to be interfered by outside noise and preprocessing process, therefore selection closes on power minimum
One group of feature vector replaces minimum value, i.e. the 5th percentile feature vector of power.
Step (1.3) the RNN-LSTM learning algorithms train location model, include the following steps：
(1.3.1) determines parameter：Initialize the weight matrix of input layer, hidden layer, output layer；
(1.3.2) propagated forward：The output valve of each neuron of forward calculation；
(1.3.3) backpropagation：Propagate packet in the direction of the error term of each neuron of backwards calculation, RNN-LSTM error terms
Include both direction：One is propagated along the direction of time, i.e., since current t moment, calculates the error term at each moment；One
It is to propagate error term upper layer；
The iteration that (1.3.4) carries out parameters weighting according to corresponding error term updates calculating.
RNN-LSTM replaces the Advanced Edition of the RNN of conventional network elements using LSTM cells.The basis of LSTM cells is former
Reason is to manipulate the information flow in network with different types of door.By door, LSTM cells can decide when
It should remember input information, when should forget the information and when should export the information.Therefore it is one
Kind can protect stored complicated and exquisite network element RNN-LSTM for a long time.It can solve to disappear or explode due to gradient
Caused short cycle Dependence Problem, to realize the effect of long-term memory.
The calculating of step (2.2) the room discrimination：It is the room mark of the result and true environment that are exported according to model
Note is compared, and can calculate the discrimination p in room；
Wherein, yiIt indicates to mark using the calculated room of model,Indicate the room label under true environment,
Expression is worked asIts value is 1；Otherwise its value is 0.
The present invention is based on the room-level localization method that smart mobile phone room background phonoreception is known, this method is other without disposing in advance
Infrastructure, it is only necessary to acquire room background sound using smart mobile phone, the 5th percentile power of extraction is as background sound fingerprint characteristic.
This feature extracting method calculates simply relative to feature extracting methods such as MFCC, passes through RNN-LSTM deep learning algorithms and trains
The Model Identification rate obtained is high, compares conventional model performance and has and is largely promoted, and is more suitable for room background acoustic fix ranging field
Jing Zhong.
Description of the drawings
Fig. 1 is to be acquired and training process block diagram under indoor room background sound ray in localization method of the present invention；
Fig. 2 is position fixing process block diagram on background sound ray in place to be positioned in localization method of the present invention.
Specific implementation mode
The content of present invention is further described below in conjunction with the accompanying drawings, but is not limitation of the invention.
Referring to Fig.1-2, the present invention is based on the room-level localization methods that smart mobile phone room background phonoreception is known, including walk as follows
Suddenly：
(1) acquisition and training under indoor room background sound ray
(1.1) indoor room background sound and feature extraction are acquired：
Using enough room background sound data are acquired under smart mobile phone line, background acoustic feature extraction is carried out, passes through the 5th
Percentile power draw goes out background sound fingerprint；
(1.2) background sound fingerprint base is built：
Background sound fingerprint is collectively formed into room background sound fingerprint base with the room label information manually marked；
(1.3) training process：
After constructing background sound fingerprint base, as training set data, trained by RNN-LSTM deep learning algorithms
Go out to be suitable for the location model under background sound indoor positioning scene, this model there need to be higher generalization ability, can preferably reflect
The feature of entire sample space；
(2) it is positioned on background sound ray in place to be positioned
(2.1) background sound test set finger print data in place to be positioned is obtained：
Place background sound to be positioned in studio carries out the 5th percentage power draw, and the background sound fingerprint obtained is as survey
Examination collection data；
(2.2) it by the background sound location model of training under test set Data In-Line, after input, exports to mark for room and believe
Breath, the discrimination that can calculate room is compared by the room information with true environment, identifies that room-level can be realized in room
Positioning.
Step (1.1) the 5th percentile power draw goes out background sound fingerprint, includes the following steps：
(1.1.1) carries out framing windowing operation to the original audio sequence of acquisition, obtains the background acoustical signal of short-term stationarity,
Window function is：
Each frame audio signal after framing adding window is done FFT transform by (1.1.2), retain FFT transform preceding two/
One data, and it is multiplied by its conjugation, power spectrum can be found out；
FFT transform formula is：
(1.1.3) gives up the audio signal that frequency is more than 7kHz；
(1.1.4) is ranked up remaining data by watt level；
5th percentage of (1.1.5) extraction power arranges and takes logarithm, obtains background sound fingerprint.
The 5th percentage power draw of carry out described in step (2.1), method are identical as step (1.1).
Step (1.3) the RNN-LSTM learning algorithms train location model, include the following steps：
(1.3.1) determines parameter：Initialize the weight matrix of input layer, hidden layer, output layer；
(1.3.2) propagated forward：The output valve of each neuron of forward calculation；
(1.3.3) backpropagation：Propagate packet in the direction of the error term of each neuron of backwards calculation, RNN-LSTM error terms
Include both direction：One is propagated along the direction of time, i.e., since current t moment, calculates the error term at each moment；One
It is to propagate error term upper layer；
The iteration that (1.3.4) carries out parameters weighting according to corresponding error term updates calculating.
The calculating of step (2.2) the room discrimination：It is the room mark of the result and true environment that are exported according to model
Note is compared, and can calculate the discrimination p in room；
Wherein, yiIt indicates to mark using the calculated room of model,Indicate the room label under true environment,
Expression is worked asIts value is 1；Otherwise its value is 0.
The present invention enrolls indoor environment sound using smart mobile phone, and the 5th percentile power by extracting sound power is used as
Acoustic feature imports RNN-LSTM learning algorithms, trains background sound location model in given chamber, relatively traditional similar acoustic signature room
For interior positioning system, reaches 90% or more using discrimination of the method for the present invention in 15 rooms, do not only reach room-level
The requirement of positioning, and improve room discrimination.
Claims (4)
1. based on the room-level localization method that smart mobile phone room background phonoreception is known, including being adopted under (1) indoor room background sound ray
Collection and training, and two stages are positioned on (2) background sound ray in place to be positioned, it is characterised in that：
Acquisition and training, specific method include the following steps under stage (1) the indoor room background sound ray:
(1.1) indoor room background sound and feature extraction are acquired：
Using enough room background sound data are acquired under smart mobile phone line, background acoustic feature extraction is carried out, the 5th percentage is passed through
Position power draw goes out background sound fingerprint；
(1.2) background sound fingerprint base is built：
Background sound fingerprint is collectively formed into room background sound fingerprint base with the room label information manually marked；
(1.3) training process：
After constructing background sound fingerprint base, as training set data, trained by RNN-LSTM deep learning algorithms suitable
For the location model under background sound indoor positioning scene；
It is positioned on stage (2) the background sound ray in place to be positioned, specific method includes the following steps:
(2.1) background sound test set finger print data in place to be positioned is obtained：
Place background sound to be positioned in studio carries out the 5th percentage power draw, and the background sound fingerprint obtained is as test set
Data；
(2.2) will under test set Data In-Line training background sound location model, after input, export as room label information,
The discrimination in room can be calculated by being compared by the room information with true environment, and it is fixed to identify that room-level can be realized in room
Position.
2. the room-level localization method according to claim 1 known based on smart mobile phone room background phonoreception, feature are existed
In：Step (1.1) the 5th percentile power draw goes out background sound fingerprint, includes the following steps：The original of (1.1.1) to acquisition
Beginning tonic train carries out framing windowing operation, obtains the background acoustical signal of short-term stationarity, window function is：
Each frame audio signal after framing adding window is done FFT transform by (1.1.2), retains the preceding half number of FFT transform
According to, and it is multiplied by its conjugation, power spectrum can be found out；
FFT transform formula is：
(1.1.3) gives up the audio signal that frequency is more than 7kHz；
(1.1.4) is ranked up remaining data by watt level；
5th percentage of (1.1.5) extraction power arranges and takes logarithm, obtains background sound fingerprint.
3. the room-level localization method according to claim 1 known based on smart mobile phone room background phonoreception, feature are existed
In：Step (1.3) the RNN-LSTM learning algorithms train location model, include the following steps：
(1.3.1) determines parameter：Initialize the weight matrix of input layer, hidden layer, output layer；
(1.3.2) propagated forward：The output valve of each neuron of forward calculation；
(1.3.3) backpropagation：The error term of each neuron of backwards calculation, it includes two that the direction of RNN-LSTM error terms, which is propagated,
A direction：One is propagated along the direction of time, i.e., since current t moment, calculates the error term at each moment；One be by
Error term upper layer is propagated；
The iteration that (1.3.4) carries out parameters weighting according to corresponding error term updates calculating.
4. the room-level localization method according to claim 1 known based on smart mobile phone room background phonoreception, feature are existed
In：The calculating of step (2.2) the room discrimination：It is the room label of the result and true environment that are exported according to location model
It is compared, the discrimination p in room can be calculated；
Wherein, yiIt indicates to mark using the calculated room of model,Indicate the room label under true environment,Expression is worked asIts value is 1；Otherwise its value is 0.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810560130.9A CN108810838A (en) | 2018-06-03 | 2018-06-03 | The room-level localization method known based on smart mobile phone room background phonoreception |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810560130.9A CN108810838A (en) | 2018-06-03 | 2018-06-03 | The room-level localization method known based on smart mobile phone room background phonoreception |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108810838A true CN108810838A (en) | 2018-11-13 |
Family
ID=64090138
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810560130.9A Pending CN108810838A (en) | 2018-06-03 | 2018-06-03 | The room-level localization method known based on smart mobile phone room background phonoreception |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108810838A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109547936A (en) * | 2018-12-29 | 2019-03-29 | 桂林电子科技大学 | Indoor orientation method based on Wi-Fi signal and environmental background sound |
| CN110333484A (en) * | 2019-07-15 | 2019-10-15 | 桂林电子科技大学 | The room area grade localization method with analysis is known based on environmental background phonoreception |
| CN111415678A (en) * | 2019-01-07 | 2020-07-14 | 意法半导体公司 | Open or closed space environment classification for mobile or wearable devices |
| CN112040408A (en) * | 2020-08-14 | 2020-12-04 | 山东大学 | Multi-target accurate intelligent positioning and tracking method suitable for supervision places |
| CN114339600A (en) * | 2022-01-10 | 2022-04-12 | 浙江德清知路导航科技有限公司 | Electronic equipment indoor positioning system and method based on 5G signal and sound wave signal |
| US20220317272A1 (en) * | 2021-03-31 | 2022-10-06 | At&T Intellectual Property I, L.P. | Using Scent Fingerprints and Sound Fingerprints for Location and Proximity Determinations |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020097882A1 (en) * | 2000-11-29 | 2002-07-25 | Greenberg Jeffry Allen | Method and implementation for detecting and characterizing audible transients in noise |
| CN105976827A (en) * | 2016-05-26 | 2016-09-28 | 南京邮电大学 | Integrated-learning-based indoor sound source positioning method |
| CN106535134A (en) * | 2016-11-22 | 2017-03-22 | 上海斐讯数据通信技术有限公司 | Multi-room locating method based on wifi and server |
| CN107703486A (en) * | 2017-08-23 | 2018-02-16 | 南京邮电大学 | A kind of auditory localization algorithm based on convolutional neural networks CNN |
-
2018
- 2018-06-03 CN CN201810560130.9A patent/CN108810838A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020097882A1 (en) * | 2000-11-29 | 2002-07-25 | Greenberg Jeffry Allen | Method and implementation for detecting and characterizing audible transients in noise |
| CN105976827A (en) * | 2016-05-26 | 2016-09-28 | 南京邮电大学 | Integrated-learning-based indoor sound source positioning method |
| CN106535134A (en) * | 2016-11-22 | 2017-03-22 | 上海斐讯数据通信技术有限公司 | Multi-room locating method based on wifi and server |
| CN107703486A (en) * | 2017-08-23 | 2018-02-16 | 南京邮电大学 | A kind of auditory localization algorithm based on convolutional neural networks CNN |
Non-Patent Citations (2)
| Title |
|---|
| TARZIA S P ET AL: "Indoor Localization without Infrastructure Using the Acoustic Background Spectrum", 《INTENATIONAL CONFERENCE ON MOBILE SYSTEMS,APPLICATIONS AND SERVICES,ACM》 * |
| 陈文婧: "基于环境感知的智能手机室内定位系统的设计和实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109547936A (en) * | 2018-12-29 | 2019-03-29 | 桂林电子科技大学 | Indoor orientation method based on Wi-Fi signal and environmental background sound |
| CN111415678A (en) * | 2019-01-07 | 2020-07-14 | 意法半导体公司 | Open or closed space environment classification for mobile or wearable devices |
| CN111415678B (en) * | 2019-01-07 | 2024-02-27 | 意法半导体公司 | Classifying open or closed space environments for mobile or wearable devices |
| CN110333484A (en) * | 2019-07-15 | 2019-10-15 | 桂林电子科技大学 | The room area grade localization method with analysis is known based on environmental background phonoreception |
| CN110333484B (en) * | 2019-07-15 | 2021-04-13 | 桂林电子科技大学 | Indoor area level positioning method based on environmental background sound perception and analysis |
| CN112040408A (en) * | 2020-08-14 | 2020-12-04 | 山东大学 | Multi-target accurate intelligent positioning and tracking method suitable for supervision places |
| CN112040408B (en) * | 2020-08-14 | 2021-08-03 | 山东大学 | Multi-target accurate intelligent positioning and tracking method suitable for supervision places |
| US20220317272A1 (en) * | 2021-03-31 | 2022-10-06 | At&T Intellectual Property I, L.P. | Using Scent Fingerprints and Sound Fingerprints for Location and Proximity Determinations |
| CN114339600A (en) * | 2022-01-10 | 2022-04-12 | 浙江德清知路导航科技有限公司 | Electronic equipment indoor positioning system and method based on 5G signal and sound wave signal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108810838A (en) | The room-level localization method known based on smart mobile phone room background phonoreception | |
| CN103310789B (en) | A kind of sound event recognition method of the parallel model combination based on improving | |
| CN101023469B (en) | Digital filtering method, digital filtering equipment | |
| CN107610707A (en) | A kind of method for recognizing sound-groove and device | |
| US20190392858A1 (en) | Intelligent voice outputting method, apparatus, and intelligent computing device | |
| CN109949823A (en) | A kind of interior abnormal sound recognition methods based on DWPT-MFCC and GMM | |
| CN108630209B (en) | Marine organism identification method based on feature fusion and deep confidence network | |
| CN101923855A (en) | Test-irrelevant voice print identifying system | |
| CN101710490A (en) | Method and device for compensating noise for voice assessment | |
| CN104916289A (en) | Quick acoustic event detection method under vehicle-driving noise environment | |
| CN108922541A (en) | Multidimensional characteristic parameter method for recognizing sound-groove based on DTW and GMM model | |
| CN109192224A (en) | A kind of speech evaluating method, device, equipment and readable storage medium storing program for executing | |
| CN111724770B (en) | Audio keyword identification method for generating confrontation network based on deep convolution | |
| CN103559879A (en) | Method and device for extracting acoustic features in language identification system | |
| CN111341319B (en) | Audio scene identification method and system based on local texture features | |
| CN108597505A (en) | Audio recognition method, device and terminal device | |
| CN108520753A (en) | Voice lie detection method based on the two-way length of convolution memory network in short-term | |
| CN108877809A (en) | A kind of speaker's audio recognition method and device | |
| CN109767760A (en) | Far field audio recognition method based on the study of the multiple target of amplitude and phase information | |
| CN107202559B (en) | Object identification method based on indoor acoustic channel disturbance analysis | |
| CN111640411A (en) | Audio synthesis method, device and computer readable storage medium | |
| CN105825857A (en) | Voiceprint-recognition-based method for assisting deaf patient in determining sound type | |
| CN109884591A (en) | A kind of multi-rotor unmanned aerial vehicle acoustical signal Enhancement Method based on microphone array | |
| CN102426837B (en) | Robustness method used for voice recognition on mobile equipment during agricultural field data acquisition | |
| CN106019230B (en) | A kind of sound localization method based on i-vector Speaker Identification |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181113 |