CN109741752A - A kind of occurrences in human life examining method and system based on speech recognition - Google Patents
A kind of occurrences in human life examining method and system based on speech recognition Download PDFInfo
- Publication number
- CN109741752A CN109741752A CN201811614093.1A CN201811614093A CN109741752A CN 109741752 A CN109741752 A CN 109741752A CN 201811614093 A CN201811614093 A CN 201811614093A CN 109741752 A CN109741752 A CN 109741752A
- Authority
- CN
- China
- Prior art keywords
- examination
- occurrences
- model
- human life
- speech recognition
- 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
Landscapes
- Machine Translation (AREA)
Abstract
The present invention provides a kind of occurrences in human life examining method and system based on speech recognition, comprising: S101, examination group ID and examination batch ID is established, to record examination；S102, receive voice signal, using syllable as recognition unit, using based on convolutional neural networks acoustic model and hidden Markov language model convert voice in real time corresponding text；S103, examination keywords database is established, full-text search is carried out to memcon according to dictionary content, counts crucial word frequency and modeling analysis.The present invention realizes the transcription job of from voice signal to text file (memcon), and then replace the working method of traditional-handwork translation, translation result accuracy rate is high, system application is strong, a large amount of human and material resources, while the foundation of analysis model are saved, appraisal result is more intuitively showed, subjectivity and the one-sidedness result of appraisal caused by artificial examination are effectively reduced, ensure the fair and just property of occurrences in human life examination.
Description
Technical field
The present invention relates to Techniques of Enterprise Management field, especially a kind of occurrences in human life examining method based on speech recognition be
System.
Background technique
Evaluation of employee system is that enterprise is used to manage the important tool of enterprise staff working condition, task mainly include with
Under several: information management is mainly responsible for management employee and employs information and essential information, while providing the function of register inquiry
Energy；Salary management is mainly responsible for the information such as wage of management standard, wage meter hair and payroll journal；Staff attendance management module, it is main
It is responsible for staff attendance typing, attendance query and attendance examination；Job rating management, the work for being mainly responsible for management employee are examined
Nuclear information, including the single generation of typing, inquiry and examination；Information inquiry, can mainly allow administrative staff to position rapidly and oneself think
The information to be searched improves working efficiency.And the performance assessment criteria that can not quantify is needed to check and rate by occurrences in human life and is talked to obtain.
It is working attitude, ability to work and work of the enterprise by reasonable evaluation means to employee that occurrences in human life examination, which is talked,
The important personnel management process that achievement is made an appraisal.Personnel evaluation the result is that the promotion (degradation) of employee, wages, welfare,
Bonus and the foundation of personnel assignment.
Traditional occurrences in human life examination is talked mainly using the form of " sound recordings, human translation are written to summarize ", assessment work
Time-consuming, and appraisal result is subjective, unilateral, with the development of the technologies such as computer science and communication, is based on speech recognition technology, such as
What discharges labor workload by information-based means, and the information-based work capacity for promoting occurrences in human life examination has been the task of top priority.
Summary of the invention
The object of the present invention is to provide a kind of occurrences in human life examining method and system based on speech recognition, it is intended to solve existing skill
In art occurrences in human life examination talk need to by it is artificial realize, time-consuming low efficiency the problem of, realize and improve translation result accuracy rate, save
A large amount of human and material resources.
To reach above-mentioned technical purpose, the present invention provides a kind of occurrences in human life examining method based on speech recognition, the side
Method the following steps are included:
S101, examination group ID and examination batch ID is established, to record examination；
S102, receive voice signal, using syllable as recognition unit, using based on convolutional neural networks acoustic model with
Hidden Markov language model converts voice to corresponding text in real time；
S103, examination keywords database is established, full-text search is carried out to memcon according to dictionary content, counts crucial word frequency
And modeling analysis.
Preferably, the acoustic model and hidden Markov language model using based on convolutional neural networks turns voice
Turning to corresponding text, specific step is as follows:
Construct convolutional neural networks model；
Acoustic model building is realized by the sound bank constructed in advance, and the voice for extracting every frame regular length is corresponding
FBank feature is trained acoustic model by convolutional neural networks model, converts speech into phonetic；
Phonetic is converted into text using hidden Markov language model.
Preferably, the convolutional neural networks model is in such a way that convolutional layer and pond layer are arranged alternately, the convolution
The activation value of layer is calculated by the following formula:
hj,kIndicate that the jth characteristic pattern in convolution output layer, k-th of neuronal activation value, s indicate the length of convolution kernel,Indicate the weighted value for corresponding to b-th of frequency band of j-th of convolution kernel, ajIt is the biasing of corresponding jth characteristic pattern, θ (x) is
Activation primitive.
Preferably, the very big pond function formula of the pond layer is as follows:
pj,mIndicate the output of pond layer, j indicates that j-th of characteristic pattern, m indicate m-th of pond band, and n is the down-sampling factor, r
It is the size in pond.
Preferably, the method also includes talking to being transcribed offline to examination, specifically:
S201, external audio file is imported；
S202, using syllable as recognition unit, by speech signal pre-processing and Feature Extraction Technology, to audio file into
Row translation；
S203, editor's adjustment translation content.
The present invention also provides a kind of occurrences in human life appraisal system based on speech recognition, the system comprises:
Examination ID establishes module, for establishing examination group ID and examination batch ID, to record examination；
Real time transcription module, for receiving voice signal, using syllable as recognition unit, using based on convolutional neural networks
Acoustic model and hidden Markov language model convert voice in real time corresponding text；
Statistical analysis module carries out full-text search to memcon according to dictionary content for establishing examination keywords database,
Count crucial word frequency and modeling analysis.
Preferably, the real time transcription module includes:
Deep learning unit, for constructing convolutional neural networks model；
Acoustic training model unit realizes acoustic model building for the sound bank by constructing in advance, it is solid to extract every frame
The corresponding fBank feature of the voice of measured length, is trained acoustic model by convolutional neural networks model, digitizes the speech into
At phonetic；
Text conversion unit, for phonetic to be converted into text using hidden Markov language model.
Preferably, the convolutional neural networks model is in such a way that convolutional layer and pond layer are arranged alternately, the convolution
The activation value of layer is calculated by the following formula:
hj,kIndicate that the jth characteristic pattern in convolution output layer, k-th of neuronal activation value, s indicate the length of convolution kernel,Indicate the weighted value for corresponding to b-th of frequency band of j-th of convolution kernel, ajIt is the biasing of corresponding jth characteristic pattern, θ (x) is
Activation primitive.
Preferably, the very big pond function formula of the pond layer is as follows:
pj,mIndicate the output of pond layer, j indicates that j-th of characteristic pattern, m indicate m-th of pond band, and n is the down-sampling factor, r
It is the size in pond.
Preferably, the system also includes offline transcription module, the offline transcription module includes:
Audio import unit, for importing external audio file；
Translation unit is used for using syllable as recognition unit, by speech signal pre-processing and Feature Extraction Technology, to sound
Frequency file is translated；
Content editing unit, for editing adjustment translation content.
The effect provided in summary of the invention is only the effect of embodiment, rather than invents all whole effects, above-mentioned
A technical solution in technical solution have the following advantages that or the utility model has the advantages that
Compared with prior art, the present invention by occurrences in human life check and rate talk work by means of speech recognition technology realize from
Voice signal to the transcription job of text file (memcon), be it is a kind of be converted to by voice signal it is computer-readable defeated
Enter, then management tool is checked and rated by the talk that computer-readable input is converted to text output, and then replaces traditional-handwork translation
Working method, translation result accuracy rate is high, and system application is strong, saves a large amount of human and material resources, speech recognition technology
Application, be that information tool is merged with the in-depth of business administration, while the foundation of analysis model, more intuitively show examination knot
Fruit, information-based examination analysis means effectively reduce subjectivity and the one-sidedness result of appraisal caused by artificial examination, ensure occurrences in human life
The fair and just property of examination.
Detailed description of the invention
Fig. 1 is a kind of occurrences in human life examining method flow chart based on speech recognition provided in the embodiment of the present invention；
Fig. 2 is a kind of offline dubbing method flow chart provided in the embodiment of the present invention；
Fig. 3 is a kind of occurrences in human life appraisal system structural block diagram based on speech recognition provided in the embodiment of the present invention；
Fig. 4 is a kind of occurrences in human life appraisal system structure based on offline transcription speech recognition provided in the embodiment of the present invention
Block diagram.
Specific embodiment
In order to clearly illustrate the technical characterstic of this programme, below by specific embodiment, and its attached drawing is combined, to this
Invention is described in detail.Following disclosure provides many different embodiments or example is used to realize different knots of the invention
Structure.In order to simplify disclosure of the invention, hereinafter the component of specific examples and setting are described.In addition, the present invention can be with
Repeat reference numerals and/or letter in different examples.This repetition is that for purposes of simplicity and clarity, itself is not indicated
Relationship between various embodiments and/or setting is discussed.It should be noted that illustrated component is not necessarily to scale in the accompanying drawings
It draws.Present invention omits the descriptions to known assemblies and treatment technology and process to avoid the present invention is unnecessarily limiting.
It is provided for the embodiments of the invention a kind of occurrences in human life examining method based on speech recognition with reference to the accompanying drawing and is
System is described in detail.
As shown in Figure 1, the embodiment of the invention discloses a kind of occurrences in human life examining method based on speech recognition, the method packet
Include following steps:
S101, examination group ID and examination batch ID is established, to record examination；
S102, receive voice signal, using syllable as recognition unit, using based on convolutional neural networks acoustic model with
Hidden Markov language model converts voice to corresponding text in real time；
S103, examination keywords database is established, full-text search is carried out to memcon according to dictionary content, counts crucial word frequency
And modeling analysis.
Before occurrences in human life examination is talked and started, according to specific talk examination system, establish examination batch, be arranged appraiser and by
Appraiser determines examination range, is grouped to appraisee that reasonable distribution checks and rates task, guarantees talk assessment work on time
It completes, establishes examination group ID and examination batch ID, assessment work is managed collectively.
When checking and rating talk, transcribed by real time transcription technology in recording.Real-time reception voice signal in conversation on course,
Using syllable as recognition unit, the real time translation to voice signal is completed by speech recognition technology.One section of sound is resolved to
Smaller voice unit (VU) is converted into then by acoustic model and Hidden Markov language model based on convolutional neural networks
Corresponding text, is handled by language model, and optimal text results can be obtained.
It is as follows for the process of examination conversational speech identification:
Basic framework using Hidden Markov language model as Acoustic Modeling.
Convolutional neural networks model is constructed, in such a way that convolutional layer and pond layer are arranged alternately, the convolutional layer is by more
A characteristic pattern forms, and all neurons on characteristic pattern share a convolution kernel.The activation value of convolutional layer passes through following formula meter
It calculates:
hj,kIndicate that the jth characteristic pattern in convolution output layer, k-th of neuronal activation value, s indicate the length of convolution kernel,Indicate the weighted value for corresponding to b-th of frequency band of j-th of convolution kernel, ajIt is the biasing of corresponding jth characteristic pattern, θ (x) is
Activation primitive.
For pond layer, to pass through the expression of the low resolution of down-sampled method calculating convolutional layer activation value, very big pond
The formula for changing function is as follows:
pj,mIndicate the output of pond layer, j indicates that j-th of characteristic pattern, m indicate m-th of pond band, and n is the down-sampling factor, r
It is the size in pond.
Acoustic model building is realized by the sound bank constructed in advance, and the voice for extracting every frame regular length is corresponding
FBank feature is trained acoustic model by convolutional neural networks model, converts speech into phonetic, finally using hidden
Phonetic is converted into text by markov language model.
The reality to voice signal is completed by acoustic model based on convolutional neural networks and Hidden Markov language model
When translate, realize the real-time corresponding of audio file and text file, and realize the editable function of text file (memcon)
Can, user can be edited in real time for the content currently transcribed.It records after completing, system saves current recording text
Part and transcrypt content can carry out manual modification to conversation content in transcription, while can check the basic letter of appraisee
Breath and appraisee correspond to the relevant issues in post.
Former phonic signal character can be made to enhance by using the acoustic model based on convolutional neural networks, and can reduce and make an uproar
Sound using speech signal spec-trum local correlations principle, carries out sub-sample to feature in pond layer, can to Data Dimensionality Reduction and
Retain useful information.
The embodiment of the present invention also has offline functional transcription other than real time transcription.External audio file can be with introgressive line
System, using syllable as recognition unit, by speech signal pre-processing and Feature Extraction Technology, translates audio file, turns over
Translate the adjustment of content editable.For the recording file in system, play time section can be arbitrarily pulled, compiling again for transcrypt content is facilitated
Volume, as shown in Figure 2.
After the completion of transcription, examination keyword thesaurus need to be established, according to dictionary according to the crucial performance assessment criteria of appraisee
Content carries out full-text search to memcon, counts crucial word frequency modeling analysis, and visual pattern shows appraisal result.By right
Conversation content is for statistical analysis, carries out crucial word frequency statistics to post key index, can be according to statistical result to different indexs
It gives a mark, according to the weight proportion of marking situation and index, calculates score, built according to index score and total score
Mould, and then show appraisal result.
The embodiment of the present invention is realized by means of speech recognition technology from voice signal by checking and rating talk work to occurrences in human life
To the transcription job of text file (memcon), it is that one kind by voice signal is converted to computer-readable input, then by counting
Calculation machine can read input be converted to text output talk examination management tool, and then replace traditional-handwork translation working method,
Translation result accuracy rate is high, and system application is strong, saves a large amount of human and material resources, and the application of speech recognition technology is
Information tool is merged with the in-depth of business administration, while the foundation of analysis model, more intuitively shows appraisal result, information-based
Examination analysis means, effectively reduce subjectivity and the one-sidedness result of appraisal caused by artificial examination, ensure the public affairs of occurrences in human life examination
Flat fairness.
As shown in figure 3, the embodiment of the invention also discloses a kind of occurrences in human life appraisal system based on speech recognition, the system
Include:
Examination ID establishes module, for establishing examination group ID and examination batch ID, to record examination；
Real time transcription module, for receiving voice signal, using syllable as recognition unit, using based on convolutional neural networks
Acoustic model and hidden Markov language model convert voice in real time corresponding text；
Statistical analysis module carries out full-text search to memcon according to dictionary content for establishing examination keywords database,
Count crucial word frequency and modeling analysis.
The real time transcription module includes:
Deep learning unit, for constructing convolutional neural networks model；
Acoustic training model unit realizes acoustic model building for the sound bank by constructing in advance, it is solid to extract every frame
The corresponding fBank feature of the voice of measured length, is trained acoustic model by convolutional neural networks model, digitizes the speech into
At phonetic；
Text conversion unit, for phonetic to be converted into text using hidden Markov language model.
The reality to voice signal is completed by acoustic model based on convolutional neural networks and Hidden Markov language model
When translate, realize the real-time corresponding of audio file and text file, and realize the editable function of text file (memcon)
Can, user can be edited in real time for the content currently transcribed.It records after completing, system saves current recording text
Part and transcrypt content can carry out manual modification to conversation content in transcription, while can check the basic letter of appraisee
Breath and appraisee correspond to the relevant issues in post.
The convolutional neural networks model is in such a way that convolutional layer and pond layer are arranged alternately, the activation of the convolutional layer
Value is calculated by the following formula:
hj,kIndicate that the jth characteristic pattern in convolution output layer, k-th of neuronal activation value, s indicate the length of convolution kernel,Indicate the weighted value for corresponding to b-th of frequency band of j-th of convolution kernel, ajIt is the biasing of corresponding jth characteristic pattern, θ (x) is
Activation primitive.
The very big pond function formula of the pond layer is as follows:
pj,mIndicate the output of pond layer, j indicates that j-th of characteristic pattern, m indicate m-th of pond band, and n is the down-sampling factor, r
It is the size in pond.
Acoustic model building is realized by the sound bank constructed in advance, and the voice for extracting every frame regular length is corresponding
FBank feature is trained acoustic model by convolutional neural networks model, converts speech into phonetic, finally using hidden
Phonetic is converted into text by markov language model.
The system also includes offline transcription modules, as shown in figure 4, the offline transcription module includes:
Audio import unit, for importing external audio file；
Translation unit is used for using syllable as recognition unit, by speech signal pre-processing and Feature Extraction Technology, to sound
Frequency file is translated；
Content editing unit, for editing adjustment translation content.
After transcription is completed to generate text, according to the crucial performance assessment criteria of appraisee, examination keyword thesaurus, root are established
Full-text search is carried out to memcon according to dictionary content, counts crucial word frequency modeling analysis, visual pattern shows appraisal result.
By for statistical analysis to conversation content, crucial word frequency statistics are carried out to post key index, it can be according to statistical result to not
Give a mark with index, according to the weight proportion of marking situation and index, calculate score, according to index score and total score into
Row modeling, and then show appraisal result.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.
Claims (10)
1. a kind of occurrences in human life examining method based on speech recognition, which is characterized in that the described method comprises the following steps:
S101, examination group ID and examination batch ID is established, to record examination；
S102, receive voice signal, using syllable as recognition unit, using based on convolutional neural networks acoustic model and hidden horse
Er Kefu language model converts voice to corresponding text in real time；
S103, examination keywords database is established, full-text search is carried out to memcon according to dictionary content, counted crucial word frequency and build
Mould analysis.
2. a kind of occurrences in human life examining method based on speech recognition according to claim 1, which is characterized in that described to use base
Corresponding text detailed process is converted such as voice in the acoustic model and hidden Markov language model of convolutional neural networks
Under:
Construct convolutional neural networks model；
Acoustic model building is realized by the sound bank constructed in advance, and the corresponding fBank of voice for extracting every frame regular length is special
Sign, is trained acoustic model by convolutional neural networks model, converts speech into phonetic；
Phonetic is converted into text using hidden Markov language model.
3. a kind of occurrences in human life examining method based on speech recognition according to claim 1 or 2, which is characterized in that the volume
For product neural network model in such a way that convolutional layer and pond layer are arranged alternately, the activation value of the convolutional layer passes through following formula
It calculates:
hj,kIndicate that the jth characteristic pattern in convolution output layer, k-th of neuronal activation value, s indicate the length of convolution kernel,
Indicate the weighted value for corresponding to b-th of frequency band of j-th of convolution kernel, ajIt is the biasing of corresponding jth characteristic pattern, θ (x) is activation
Function.
4. a kind of occurrences in human life examining method based on speech recognition according to claim 3, which is characterized in that the pond layer
Very big pond function formula it is as follows:
pj,mIndicate the output of pond layer, j indicates that j-th of characteristic pattern, m indicate m-th of pond band, and n is the down-sampling factor, and r is pond
Size.
5. a kind of occurrences in human life examining method based on speech recognition according to claim 1, which is characterized in that the method is also
It is transcribed offline including talking to examination, specifically:
S201, external audio file is imported；
S202, using syllable as recognition unit, by speech signal pre-processing and Feature Extraction Technology, audio file is turned over
It translates；
S203, editor's adjustment translation content.
6. a kind of occurrences in human life appraisal system based on speech recognition, which is characterized in that the system comprises:
Examination ID establishes module, for establishing examination group ID and examination batch ID, to record examination；
Real time transcription module, for receiving voice signal, using syllable as recognition unit, using the sound based on convolutional neural networks
It learns model and hidden Markov language model converts voice to corresponding text in real time；
Statistical analysis module carries out full-text search, statistics to memcon according to dictionary content for establishing examination keywords database
Crucial word frequency and modeling analysis.
7. a kind of occurrences in human life appraisal system based on speech recognition according to claim 6, which is characterized in that described to turn in real time
Recording module includes:
Deep learning unit, for constructing convolutional neural networks model；
Acoustic training model unit realizes acoustic model building for the sound bank by constructing in advance, extracts the fixed length of every frame
The corresponding fBank feature of the voice of degree, is trained acoustic model by convolutional neural networks model, converts speech into spelling
Sound；
Text conversion unit, for phonetic to be converted into text using hidden Markov language model.
8. a kind of occurrences in human life appraisal system based on speech recognition according to claim 6 or 7, which is characterized in that the volume
For product neural network model in such a way that convolutional layer and pond layer are arranged alternately, the activation value of the convolutional layer passes through following formula
It calculates:
hj,kIndicate that the jth characteristic pattern in convolution output layer, k-th of neuronal activation value, s indicate the length of convolution kernel,
Indicate the weighted value for corresponding to b-th of frequency band of j-th of convolution kernel, ajIt is the biasing of corresponding jth characteristic pattern, θ (x) is activation
Function.
9. a kind of occurrences in human life appraisal system based on speech recognition according to claim 8, which is characterized in that the pond layer
Very big pond function formula it is as follows:
pj,mIndicate the output of pond layer, j indicates that j-th of characteristic pattern, m indicate m-th of pond band, and n is the down-sampling factor, and r is pond
Size.
10. a kind of occurrences in human life appraisal system based on speech recognition according to claim 6, which is characterized in that the system
It further include offline transcription module, the offline transcription module includes:
Audio import unit, for importing external audio file；
Translation unit is used for using syllable as recognition unit, by speech signal pre-processing and Feature Extraction Technology, to audio text
Part is translated；
Content editing unit, for editing adjustment translation content.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811614093.1A CN109741752A (en) | 2018-12-27 | 2018-12-27 | A kind of occurrences in human life examining method and system based on speech recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811614093.1A CN109741752A (en) | 2018-12-27 | 2018-12-27 | A kind of occurrences in human life examining method and system based on speech recognition |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109741752A true CN109741752A (en) | 2019-05-10 |
Family
ID=66360225
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811614093.1A Pending CN109741752A (en) | 2018-12-27 | 2018-12-27 | A kind of occurrences in human life examining method and system based on speech recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109741752A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110992949A (en) * | 2019-11-29 | 2020-04-10 | 秒针信息技术有限公司 | Performance assessment method and device based on voice recognition and readable storage medium |
| CN111695298A (en) * | 2020-06-03 | 2020-09-22 | 重庆邮电大学 | Power system power flow simulation interaction method based on pandapplicator and voice recognition |
| CN112308379A (en) * | 2020-09-30 | 2021-02-02 | 音数汇元(上海)智能科技有限公司 | Service order evaluation method, device, equipment and storage medium for home care |
| CN112435664A (en) * | 2020-11-11 | 2021-03-02 | 郑州捷安高科股份有限公司 | Evaluation system and method based on voice recognition and electronic equipment |
| CN112466285A (en) * | 2020-12-23 | 2021-03-09 | 北京百度网讯科技有限公司 | Offline voice recognition method and device, electronic equipment and storage medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101460995A (en) * | 2006-02-07 | 2009-06-17 | 日本电气株式会社 | Monitoring device, evaluation data selection device, reception person evaluation device, and reception person evaluation system and program |
| CN104318373A (en) * | 2014-10-22 | 2015-01-28 | 沈阳化工大学 | Company personnel management system |
| CN107038220A (en) * | 2017-03-20 | 2017-08-11 | 北京光年无限科技有限公司 | Method, intelligent robot and system for generating memorandum |
| CN107368948A (en) * | 2017-06-21 | 2017-11-21 | 厦门快商通科技股份有限公司 | A kind of simulation test checking system for customer service post |
| US20180114159A1 (en) * | 2016-10-24 | 2018-04-26 | Accenture Global Solutions Limited | Task Transformation Responsive to Confidentiality Assessments |
| CN108364160A (en) * | 2017-01-26 | 2018-08-03 | 樊少霞 | A kind of intelligence personnel work processing method and processing device |
| CN108460027A (en) * | 2018-02-14 | 2018-08-28 | 广东外语外贸大学 | A kind of spoken language instant translation method and system |
| CN108564941A (en) * | 2018-03-22 | 2018-09-21 | 腾讯科技（深圳）有限公司 | Audio recognition method, device, equipment and storage medium |
| CN108921399A (en) * | 2018-06-14 | 2018-11-30 | 北京新广视通科技有限公司 | A kind of intelligence direct management system and method |
-
2018
- 2018-12-27 CN CN201811614093.1A patent/CN109741752A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101460995A (en) * | 2006-02-07 | 2009-06-17 | 日本电气株式会社 | Monitoring device, evaluation data selection device, reception person evaluation device, and reception person evaluation system and program |
| CN104318373A (en) * | 2014-10-22 | 2015-01-28 | 沈阳化工大学 | Company personnel management system |
| US20180114159A1 (en) * | 2016-10-24 | 2018-04-26 | Accenture Global Solutions Limited | Task Transformation Responsive to Confidentiality Assessments |
| CN108364160A (en) * | 2017-01-26 | 2018-08-03 | 樊少霞 | A kind of intelligence personnel work processing method and processing device |
| CN107038220A (en) * | 2017-03-20 | 2017-08-11 | 北京光年无限科技有限公司 | Method, intelligent robot and system for generating memorandum |
| CN107368948A (en) * | 2017-06-21 | 2017-11-21 | 厦门快商通科技股份有限公司 | A kind of simulation test checking system for customer service post |
| CN108460027A (en) * | 2018-02-14 | 2018-08-28 | 广东外语外贸大学 | A kind of spoken language instant translation method and system |
| CN108564941A (en) * | 2018-03-22 | 2018-09-21 | 腾讯科技（深圳）有限公司 | Audio recognition method, device, equipment and storage medium |
| CN108921399A (en) * | 2018-06-14 | 2018-11-30 | 北京新广视通科技有限公司 | A kind of intelligence direct management system and method |
Non-Patent Citations (1)
| Title |
|---|
| 张建华: "基于深度学习的语音识别应用研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110992949A (en) * | 2019-11-29 | 2020-04-10 | 秒针信息技术有限公司 | Performance assessment method and device based on voice recognition and readable storage medium |
| CN111695298A (en) * | 2020-06-03 | 2020-09-22 | 重庆邮电大学 | Power system power flow simulation interaction method based on pandapplicator and voice recognition |
| CN112308379A (en) * | 2020-09-30 | 2021-02-02 | 音数汇元(上海)智能科技有限公司 | Service order evaluation method, device, equipment and storage medium for home care |
| CN112435664A (en) * | 2020-11-11 | 2021-03-02 | 郑州捷安高科股份有限公司 | Evaluation system and method based on voice recognition and electronic equipment |
| CN112466285A (en) * | 2020-12-23 | 2021-03-09 | 北京百度网讯科技有限公司 | Offline voice recognition method and device, electronic equipment and storage medium |
| CN112466285B (en) * | 2020-12-23 | 2022-01-28 | 北京百度网讯科技有限公司 | Offline voice recognition method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109741752A (en) | A kind of occurrences in human life examining method and system based on speech recognition | |
| CN108737667B (en) | Voice quality inspection method and device, computer equipment and storage medium | |
| CN102760436B (en) | Voice lexicon screening method | |
| CN107092596B (en) | Text emotion analysis method based on attention CNNs and CCR | |
| CN110717018A (en) | Industrial equipment fault maintenance question-answering system based on knowledge graph | |
| CN109065032B (en) | External corpus speech recognition method based on deep convolutional neural network | |
| CN110942229A (en) | Service quality evaluation method and device, electronic equipment and storage medium | |
| CN110188331A (en) | Model training method, conversational system evaluation method, device, equipment and storage medium | |
| CN114116994A (en) | Welcome robot dialogue method | |
| CN110335609A (en) | A kind of air-ground communicating data analysis method and system based on speech recognition | |
| CN110457432A (en) | Interview methods of marking, device, equipment and storage medium | |
| CN107945805A (en) | A kind of intelligent across language voice identification method for transformation | |
| CN106228977A (en) | The song emotion identification method of multi-modal fusion based on degree of depth study | |
| CN109410914A (en) | A kind of Jiangxi dialect phonetic and dialect point recognition methods | |
| CN106339806A (en) | Industry holographic image constructing method and industry holographic image constructing system for enterprise information | |
| CN106128450A (en) | The bilingual method across language voice conversion and system thereof hidden in a kind of Chinese | |
| CN111400469A (en) | Intelligent generation system and method for voice question answering | |
| CN108877769B (en) | Method and device for identifying dialect type | |
| CN107093422A (en) | A kind of audio recognition method and speech recognition system | |
| CN106297769B (en) | A kind of distinctive feature extracting method applied to languages identification | |
| CN106356054A (en) | Method and system for collecting information of agricultural products based on voice recognition | |
| Nishimura et al. | Automatic n-gram language model creation from web resources | |
| CN106205635A (en) | Method of speech processing and system | |
| CN105957517A (en) | Voice data structural transformation method based on open source API and system thereof | |
| CN111090726A (en) | NLP-based electric power industry character customer service interaction method |
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 |
Application publication date: 20190510 |
|
| RJ01 | Rejection of invention patent application after publication |