CN109631931A - A kind of artificial intelligence navigator - Google Patents
A kind of artificial intelligence navigator Download PDFInfo
- Publication number
- CN109631931A CN109631931A CN201811436907.7A CN201811436907A CN109631931A CN 109631931 A CN109631931 A CN 109631931A CN 201811436907 A CN201811436907 A CN 201811436907A CN 109631931 A CN109631931 A CN 109631931A
- Authority
- CN
- China
- Prior art keywords
- frame
- voice
- module
- speech
- spectrum
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/34—Route searching; Route guidance
- G01C21/36—Input/output arrangements for on-board computers
- G01C21/3605—Destination input or retrieval
- G01C21/3608—Destination input or retrieval using speech input, e.g. using speech recognition
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Automation & Control Theory (AREA)
- General Physics & Mathematics (AREA)
- Navigation (AREA)
Abstract
The invention discloses a kind of artificial intelligence navigator, the voice messagings of the voice capture device acquisition user of the artificial Intelligent navigator；Speech recognition module is used to carry out phonetic feature analysis to the voice messaging of acquisition, is converted into corresponding text information, and be sent to path planning module；Locating module sends location information to path planning module for being positioned in real time to current vehicle；Path planning module is used to plan the driving path of current vehicle according to the road map and road-map of information and the map data base storage received, and the path planned is shown by display terminal；Voice broadcast module broadcasts the path planned.The problem of artificial Intelligent navigator, safe and reasonable is easy to use, and driver can be made directly to input by voice, solves the manual palpation operation navigator occurred in the market, can endanger driver safety when operating in vehicular motion.
Description
Technical field
The present invention relates to electronic communication technology fields, and in particular to a kind of artificial intelligence navigator.
Background technique
Automatic navigator is that one kind can help user that current location is accurately positioned, and be calculated according to set destination
Stroke guides user's row to be to the driving ancillary equipment of the instrument of destination by map denotation and voice prompting two ways
Convenient for oneself driving to go on a journey, more and more people can select the interior installation navigator at oneself at present, but existing
Navigator is manipulated by being touched with hand, and this operation can bring great harm to driver safety in vehicular motion,
And when inputting specific location, it may occur that wrong word and some rarely used words are defeated does not write not come out also causes greatly driver's navigation
Difficulty.
Summary of the invention
In view of the above-mentioned problems, the present invention provides a kind of artificial intelligence navigator.
The purpose of the present invention is realized using following technical scheme:
A kind of artificial intelligence navigator, the artificial Intelligent navigator include: voice capture device, speech analysis module, determine
Position module, map data base, path planning module, display terminal and voice broadcast module；The voice capture device: for adopting
Collect the voice messaging of user；The speech recognition module will be obtained for carrying out phonetic feature analysis to the voice messaging of acquisition
Phonetic feature be converted to corresponding text information, and be sent to the path planning module；The locating module, for working as
Vehicle in front is positioned in real time, and sends location information to the path planning module；The path planning module is used for basis
Driving path of the road map and road-map of the information received and map data base storage to current vehicle
It is planned, and the path planned is shown by the display terminal；The voice broadcast module, for planning
It is broadcasted in good path.
Preferably, the voice capture device is acoustic sensor.
Preferably, the locating module is GPS positioning module.
Preferably, which further includes the module that automatically updates connecting with the map data base, described
Automatically update module, for in the map data base road map and road-map be updated.
The invention has the benefit that the artificial Intelligent navigator, safe and reasonable is easy to use, passes through the voice being equipped with
Acquisition equipment and speech analysis module can be such that driver directly inputs by voice, solve the manual palpation behaviour occurred in the market
The problem of making navigator, driver safety can be endangered when operating in vehicular motion.
Detailed description of the invention
The present invention will be further described with reference to the accompanying drawings, but the embodiment in attached drawing is not constituted to any limit of the invention
System, for those of ordinary skill in the art, without creative efforts, can also obtain according to the following drawings
Other attached drawings.
Fig. 1 is the structural block diagram of artificial intelligence navigator in the embodiment of the present invention；
Fig. 2 is the structural block diagram of speech analysis module in the embodiment of the present invention.
Appended drawing reference: voice capture device 10；Speech analysis module 20；Locating module 30；Map data base 40；Path rule
Draw module 50；Display terminal 60；Voice broadcast module 70；Automatically update module 80；Speech detection unit 21；Speech enhancement unit
22；Feature extraction unit 23；Text information converting unit 24.
Specific embodiment
The invention will be further described with the following Examples.
Fig. 1 shows a kind of artificial intelligence navigator, which includes: voice capture device 10, voice
Analysis module 20, locating module 30, map data base 40, path planning module 50, display terminal 60 and voice broadcast module 70；
The voice capture device 10 is used to acquire the voice messaging of driver；The speech recognition module 20, for the voice to acquisition
Information carries out phonetic feature analysis, and obtained phonetic feature is converted to corresponding text information, and is sent to the path rule
Draw module；The locating module 30 for being positioned in real time to current vehicle, and sends location information to the path planning
Module 40；The path planning module 40, the highway for being stored according to the information and the map data base 40 that receive
Map and road-map plan the driving path of current vehicle, and the path planned is passed through the display terminal
60 are shown；The voice broadcast module 70, for being broadcasted to the path planned.
Preferably, the voice capture device 10 is acoustic sensor.
Preferably, the locating module 30 is GPS positioning module.
Preferably, the artificial Intelligent navigator further include connect with the map data base 40 automatically update module, institute
State and automatically update module 80, for in the map data base 40 road map and road-map be updated.
The invention has the benefit that the artificial Intelligent navigator, safe and reasonable is easy to use, passes through the voice being equipped with
Acquisition equipment 10 and speech analysis module 20 can be such that driver directly inputs by voice, solve the manual touching occurred in the market
The problem of touching operation navigator, driver safety can be endangered when operating in vehicular motion.
Preferably, referring to fig. 2, the speech analysis module 20 include: speech detection unit 21, speech enhancement unit 22,
Feature extraction unit 23 and text information converting unit 24；The speech detection unit 21, for the voice messaging to acquisition into
Row detection, to obtain voice frame fragment；The speech enhancement unit 22, for carrying out enhancing processing to the voice frame fragment；
The feature extraction unit 23, for carrying out phonetic feature analysis to enhanced voice frame fragment, to obtain phonetic feature；Institute
Text information converting unit 24 is stated, for the phonetic feature to be converted to corresponding text information, and is sent to the path
Planning module.
Preferably, the voice messaging to acquisition detects, and to obtain voice frame fragment, realization process is:
(1) framing, windowing process are carried out to the voice messaging of acquisition；
(2) Fast Fourier Transform (FFT) is carried out to each frame data obtained after framing, adding window, to obtain each frame data pair
The spectrum estimation value of amplitude frequency spectrum, phase spectrum and the noise answered；
(3) the spectrum estimation value based on the obtained corresponding amplitude frequency spectrum of each frame data, phase spectrum and noise calculates every
The Delta value of one frame data judges whether each frame is speech frame according to obtained Delta value, and specific judgment method is:
If Delta (m) < λ, m frame is noise frame, if Delta (m) >=λ, m frame is speech frame, and wherein λ is to set
Fixed threshold value traverses all frames, and the frame data for belonging to speech frame are then carried out window, superposition and inverse fast Fourier transform,
The voice frame fragment in time domain can be obtained；
Wherein, the calculation formula of Delta value are as follows:
In formula, Delta (m) is the Delta value of m frame, and m=1,2 ..., M, M is frame number；Y (m, k) is the vibration of m frame
Amplitude-frequency spectrum, D (m, k) are noise spectrum estimated value, and k indicates k-th of frequency point, meets k=1,2 ..., K.
The utility model has the advantages that the voice messaging of the driver of acquisition is carried out framing, adding window and Fourier transformation in above-described embodiment
Afterwards, according to the obtained corresponding amplitude frequency spectrum of each frame data, phase spectrum and noise spectrum estimated value, each frame is found out
Then obtained Delta value is compared by Delta value with preset threshold value, and then complete whether belong to voice to present frame
The judgement of frame not only allows for the amplitude frequency spectrum of present frame and the influence of noise spectrum estimated value, Er Qieneng solving Delta value
It is enough that the adaptive judgement for whether belonging to speech frame to each frame is realized according to the amplitude frequency spectrum and noise spectrum estimated value of each frame,
The robustness for improving the speech detection unit 221, reduce subsequent voice enhancing, feature extraction, text information conversion and
The complexity of route planning, and improve the working efficiency of the artificial Intelligent navigator route planning.
In a specific embodiment, the noise spectrum estimated value of each frame can be estimated to obtain by following processes:
(1) voice messaging based on the driver after framing, adding window, Fourier transformation obtains noise according to preceding ten frame data
The initial estimate of frequency spectrum；
(2) noise spectrum estimated value is updated using the piecewise function of lower section:
In formula, D (m, k) is the noise spectrum estimated value of m frame, and Y (m, k) is the amplitude frequency spectrum of m frame, D0For noise frequency
The initial estimate of spectrum, D (m-1, k) are the noise spectrum estimated value of (m-1) frame, α1For smoothing factor, value range is:Case A indicates that present frame is noise frame, and case B indicates that present frame is speech frame.
The utility model has the advantages that preceding ten frame of usual one section of voice signal only includes noise frame, noise is obtained using preceding ten frame data
The initial estimate of frequency spectrum is then based on to obtain the amplitude frequency of the initial estimate of noise spectrum, present frame and its previous frame
Spectrum is updated the noise spectrum estimated value of present frame, obtains the noise spectrum estimated value of updated each frame, the algorithm
The update of realization that can be adaptive to the noise spectrum estimated value of each frame improves the subsequent robust to speech frame detection
Property, be conducive to the accurate detection to voice frame fragment.
Preferably, described that enhancing processing is carried out to the voice frame fragment, specifically:
(1) framing, adding window and inverse Fourier transform are carried out to the voice frame fragment, obtain respective frame amplitude frequency spectrum,
Power spectrum, noise spectrum estimated value and noise power spectrum；
(2) amplitude frequency spectrum, power spectrum, noise spectrum estimated value and noise power spectrum based on obtained each frame calculate
The noise suppression gain factor of each frame, the wherein calculation formula of the n-th frame noise suppression gain factor are as follows:
In formula, G (n) is the noise suppression gain factor of n-th frame, and Y (n, k) is the amplitude frequency spectrum of n-th frame, and D (n, k) is the
The noise spectrum estimated value of n frame, PY(n, k) is the power spectrum of n-th frame, PD(n, k) is the noise power spectrum of n-th frame, and γ is weight
Coefficient, η are noise reduction factor；
(3) based on the noise suppression gain factor, amplitude frequency spectrum and the phase spectrum of the voice frame fragment is obtained, by going
Window, superposition and inverse fast Fourier transform, the voice frame fragment after noise reduction can be obtained.
The utility model has the advantages that in the above-described embodiments, will be passed through from the voice frame fragment in the time domain that speech detection unit 21 obtains
After crossing framing, adding window and Fourier transformation, and obtain amplitude frequency spectrum, power spectrum, noise spectrum estimated value and the noise of each frame
Power spectrum calculates the noise suppression gain factor of each frame according to the calculation formula of the customized noise suppression gain factor, will
The noise suppression gain factor of obtained each frame is multiplied with corresponding amplitude frequency spectrum, and revised amplitude frequency spectrum can be obtained.
Revised amplitude frequency spectrum and its corresponding phase spectrum are subjected to window, superposition and Fast Fourier Transform (FFT), noise reduction can be obtained
Voice frame fragment afterwards, the algorithm are modified by the amplitude frequency spectrum to each frame, can effectively inhibit the speech frame piece
Random noise in section, while enhancing non-noise part.Voice frame fragment after making denoising retains the details in voice signal
Feature, be conducive to it is subsequent path planning is accurately carried out according to voice messaging, the navigation for improving the artificial Intelligent navigator is accurate
Degree.
Finally it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than the present invention is protected
The limitation of range is protected, although explaining in detail referring to preferred embodiment to the present invention, those skilled in the art are answered
Work as understanding, it can be with modification or equivalent replacement of the technical solution of the present invention are made, without departing from the reality of technical solution of the present invention
Matter and range.
Claims (6)
1. a kind of artificial intelligence navigator characterized by comprising voice capture device, speech analysis module, locating module,
Map data base, path planning module, display terminal and voice broadcast module；
The voice capture device, for acquiring the voice messaging of user；
The speech recognition module turns obtained phonetic feature for carrying out phonetic feature analysis to the voice messaging of acquisition
It is changed to corresponding text information, and is sent to the path planning module；
The locating module for being positioned in real time to current vehicle, and sends location information to the path planning module；
The path planning module, for according to the road map of the information that receives and map data base storage and
Road-map plans the driving path of current vehicle, and the path planned is shown by the display terminal
Show；
The voice broadcast module, for being broadcasted to the path planned.
2. artificial intelligence navigator according to claim 1, which is characterized in that the voice capture device is acoustics sensor
Device.
3. artificial intelligence navigator according to claim 1, which is characterized in that the locating module is GPS positioning module.
4. artificial intelligence navigator according to claim 1, which is characterized in that further include being connect with the map data base
Automatically update module, it is described to automatically update module, for the road map and road-map in the map data base
It is updated.
5. artificial intelligence navigator according to claim 1, which is characterized in that the speech analysis module includes: voice
Detection unit, speech enhancement unit, feature extraction unit and text information converting unit；
The speech detection unit, for being detected to the voice messaging of acquisition, to obtain voice frame fragment；
The speech enhancement unit, for carrying out enhancing processing to the voice frame fragment；
The feature extraction unit, for carrying out phonetic feature analysis to enhanced voice frame fragment, to obtain phonetic feature；
The text information converting unit for the phonetic feature to be converted to corresponding text information, and is sent to described
Path planning module.
6. artificial intelligence navigator according to claim 5, which is characterized in that the voice messaging to acquisition carries out
Detection, to obtain voice frame fragment, realization process is:
(1) framing, windowing process are carried out to the voice messaging of acquisition；
(2) Fast Fourier Transform (FFT) is carried out to each frame data obtained after framing, adding window, it is corresponding to obtain each frame data
The spectrum estimation value of amplitude frequency spectrum, phase spectrum and noise；
(3) the spectrum estimation value based on the obtained corresponding amplitude frequency spectrum of each frame data, phase spectrum and noise, calculates each frame
The Delta value of data judges whether each frame is speech frame according to obtained Delta value, and specific judgment method is:
If Delta (m) < λ, m frame is noise frame, if Delta (m) >=λ, m frame is speech frame, and wherein λ is setting
Threshold value traverses all frames, and the frame data for belonging to speech frame are then carried out window, superposition and inverse fast Fourier transform
Obtain the voice frame fragment in time domain；
Wherein, the calculation formula of Delta value are as follows:
In formula, Delta (m) is the Delta value of m frame, and m=1,2 ..., M, M is frame number；Y (m, k) is the amplitude frequency of m frame
Spectrum, D (m, k) are noise spectrum estimated value, and k indicates k-th of frequency point, meets k=1,2 ..., K.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811436907.7A CN109631931A (en) | 2018-11-28 | 2018-11-28 | A kind of artificial intelligence navigator |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811436907.7A CN109631931A (en) | 2018-11-28 | 2018-11-28 | A kind of artificial intelligence navigator |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109631931A true CN109631931A (en) | 2019-04-16 |
Family
ID=66069742
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811436907.7A Withdrawn CN109631931A (en) | 2018-11-28 | 2018-11-28 | A kind of artificial intelligence navigator |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109631931A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110223429A (en) * | 2019-06-19 | 2019-09-10 | 上海应用技术大学 | Voice access control system |
-
2018
- 2018-11-28 CN CN201811436907.7A patent/CN109631931A/en not_active Withdrawn
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110223429A (en) * | 2019-06-19 | 2019-09-10 | 上海应用技术大学 | Voice access control system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11475907B2 (en) | Method and device of denoising voice signal | |
| CN104335600B (en) | The method that noise reduction mode is detected and switched in multiple microphone mobile device | |
| CN105662797B (en) | A kind of Intelligent internet of things blind-guiding stick | |
| CN105529026B (en) | Speech recognition apparatus and speech recognition method | |
| CN110648553B (en) | Site reminding method, electronic equipment and computer readable storage medium | |
| US20080069364A1 (en) | Sound signal processing method, sound signal processing apparatus and computer program | |
| CN102664006B (en) | Abnormal voice detecting method based on time-domain and frequency-domain analysis | |
| CN101916567B (en) | Speech enhancement method applied to dual-microphone system | |
| CN101154383B (en) | Method and device for noise suppression, phonetic feature extraction, speech recognition and training voice model | |
| CN104464728A (en) | Speech enhancement method based on Gaussian mixture model (GMM) noise estimation | |
| CN105640748B (en) | One kind vibration guide clothes | |
| CN101976565A (en) | Dual-microphone-based speech enhancement device and method | |
| CN102347028A (en) | Double-microphone speech enhancer and speech enhancement method thereof | |
| CN109916423A (en) | Intelligent navigation equipment and its route planning method and automatic driving vehicle | |
| CN106340292A (en) | Voice enhancement method based on continuous noise estimation | |
| CN104916289A (en) | Quick acoustic event detection method under vehicle-driving noise environment | |
| CN108597505A (en) | Audio recognition method, device and terminal device | |
| CN103109320A (en) | Noise suppression device | |
| CN101894563A (en) | Voice enhancing method | |
| CN105632512A (en) | Dual-sensor voice enhancement method based on statistics model and device | |
| KR20170054794A (en) | Apparatus and method for controlling noise in vehicle | |
| CN102436809A (en) | Network speech recognition method in English oral language machine examination system | |
| CN103632677A (en) | Method and device for processing voice signal with noise, and server | |
| EP1471501A2 (en) | Speech recognition apparatus, speech recognition method, and recording medium on which speech recognition program is computer-readable recorded | |
| CN106713633A (en) | Deaf people prompt system and method, and smart phone |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190416 |
|
| WW01 | Invention patent application withdrawn after publication |