CN109068178A - A kind of video broadcasting method and player - Google Patents
A kind of video broadcasting method and player Download PDFInfo
- Publication number
- CN109068178A CN109068178A CN201811054952.6A CN201811054952A CN109068178A CN 109068178 A CN109068178 A CN 109068178A CN 201811054952 A CN201811054952 A CN 201811054952A CN 109068178 A CN109068178 A CN 109068178A
- Authority
- CN
- China
- Prior art keywords
- video
- user
- module
- knowledge point
- player
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/443—OS processes, e.g. booting an STB, implementing a Java virtual machine in an STB or power management in an STB
- H04N21/4438—Window management, e.g. event handling following interaction with the user interface
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/433—Content storage operation, e.g. storage operation in response to a pause request, caching operations
- H04N21/4331—Caching operations, e.g. of an advertisement for later insertion during playback
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/435—Processing of additional data, e.g. decrypting of additional data, reconstructing software from modules extracted from the transport stream
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/439—Processing of audio elementary streams
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/44—Processing of video elementary streams, e.g. splicing a video clip retrieved from local storage with an incoming video stream, rendering scenes according to MPEG-4 scene graphs
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/47—End-user applications
- H04N21/472—End-user interface for requesting content, additional data or services; End-user interface for interacting with content, e.g. for content reservation or setting reminders, for requesting event notification, for manipulating displayed content
- H04N21/47214—End-user interface for requesting content, additional data or services; End-user interface for interacting with content, e.g. for content reservation or setting reminders, for requesting event notification, for manipulating displayed content for content reservation or setting reminders; for requesting event notification, e.g. of sport results or stock market
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/47—End-user applications
- H04N21/478—Supplemental services, e.g. displaying phone caller identification, shopping application
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/47—End-user applications
- H04N21/485—End-user interface for client configuration
- H04N21/4858—End-user interface for client configuration for modifying screen layout parameters, e.g. fonts, size of the windows
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/83—Generation or processing of protective or descriptive data associated with content; Content structuring
- H04N21/845—Structuring of content, e.g. decomposing content into time segments
- H04N21/8456—Structuring of content, e.g. decomposing content into time segments by decomposing the content in the time domain, e.g. in time segments
Abstract
A kind of video broadcasting method including recording the play operation of user, and constructs the operation note of each user；The Automobile driving of the personalized planning and/or optimization user of user's play operation during video playing is assisted according to operation note.Player and playback method provided by the invention pass through the video semanteme of knowledge point or the machine learning based on a large number of users operation note, active user assesses the understanding of video content, the intelligence for realizing user is reviewed, is intelligently reviewed, the management of intelligent recommendation and user's attention, and the learning efficiency and quality of user are improved.
Description
Technical field
The invention belongs to computer fields, are related to a kind of video broadcasting method, in particular to one kind admires class video playing side
Method and player.
Background technique
Now, it has been emerged in large numbers on internet a large amount of brief instructional video (admiring class video), user (learner) watches religion
It learns video and is mainly divided to two class broadcast modes.
One kind is to play instructional video online by using (IE, 360, Google, red fox etc.) browser.Another kind of is to use
(learner) is locally downloading by instructional video at family, then by playout software (such as storm video, Real-player, pps,
The players such as qq is audio-visual, Baidu is audio-visual) play instructional video.
The above two classes broadcast mode can realize basic playing function (broadcasting, pause, F.F., retrogressing, play record,
Amplification, diminution etc.).For user (learner), can exist during watching instructional video by two class modes above
Four problems:
1. study needs frequently to look back when arriving the difficult point of course；2. review certain course, need frequently to drag scroll bar
The knowledge point for needing to review can be found；User is needed to carry out a large amount of search drag operation 3. finding associated knowledge point；4.
It is estimated since the human-subject test of user itself can not make video content, thus ignores important content in learning process, this
When player cannot make effective prompting to important knowledge point.
Based on four problems above, it is desirable to a player according to video content and user's attention is designed, and
And intelligence auxiliary user (learner) study, to improve learning efficiency.
Summary of the invention
To solve the above problems, one aspect of the present invention provides a kind of playback method, the play operation including recording user,
And construct the operation note of each user；The personalized planning of user's play operation and/or optimization is assisted to use according to operation note
Automobile driving of the family during video playing.
In an embodiment of the present invention, the play operation includes: the exhibition of the layout of broadcast window, broadcast window content
Existing, the control of playback progress, the control of broadcasting speed.
In an embodiment of the present invention, the layout of the broadcast window includes quantity, size, position, the shape of broadcast window
The control of shape.
In an embodiment of the present invention, the control of the playback progress includes whole story position, the play time that window plays
Control.
In an embodiment of the present invention, the broadcast window content show including automatically generate and/or show with currently
The segment of the relevant video of broadcasting content, or with the associated local video of the video or network video.
In an embodiment of the present invention, the foundation of the personalized planning for assisting user's play operation includes at least the use
One of the history play operation at family, the play operation of same video other users, video semanteme of the video.Further
, it is described assist user's play operation planning foundation priority be followed successively by from high to low user history play operation,
The semantic analysis of the play operation of other users, the video in same video.
In an embodiment of the present invention, Automobile driving of the optimization user during video playing includes: that basis is worked as
The content of preceding video clips and the segment carry out the prompting or triggering use of different modes in the significance level of entire video to user
The planning of family play operation.
Another aspect of the present invention provides a kind of player, including playing module, control module, video cache module, institute
It states playing module and plays out that the layout of window, the showing of windows content, progress, broadcasting speed are controlled to video file,
It is characterized in that,
The player further includes record data library, AI interface, knowledge point presentation module, the control module difference
With playing module, video cache module, AI interface, knowledge point present module connect, be configured to respond to video cache module,
The request drive control module that module is presented in AI interface, knowledge point controls currently playing video；
The record data library is configured as the play operation of record user；
The AI interface is configured as the interface of connection player and AI server, in response to the request of user, calls phase
The AI service answered；
The knowledge point is presented module and is configured to respond to the request of user and calls control module, will with it is currently playing
The video clips of video response are presented to user.
In an embodiment of the present invention, the AI service includes that will work as forward sight according to the operation note of knowledge mapping or user
Frequency is divided into multiple segments, recommends associated video, according in currently playing video from trend user according to currently playing video
Appearance reminds user's attention, is known according to currently playing video clips to user's push is relevant to current video segment
Know point examination paper.
In an embodiment of the present invention, the player includes that mould is recommended in the automatic division module in knowledge point, knowledge point automatically
Module is investigated in block, user's attention reminding module, knowledge point,
Current video is divided into more by the automatic division module in knowledge point according to the operation note of knowledge mapping or user
A segment；
The automatic recommending module in knowledge point recommends associated video from trend user according to currently playing video；
User's attention reminding module reminds user's attention according to the content of currently playing video；
It is related to current video segment to user's push according to currently playing video clips that module is investigated in the knowledge point
Knowledge point examination paper.
In an embodiment of the present invention, the player further includes video cache module, and the video cache module is to working as
The video of preceding broadcasting is cached.
In an embodiment of the present invention, the player includes a main broadcast window, at least one is from broadcast window, institute
Main broadcast window is stated for playing the video file of user's request；
It is described from broadcast window be used for play user request video file video clips or with currently playing window phase
The video file of pass.
Further, the main broadcast window or current video can be broadcast by user or AI interface module from broadcast window
The layout for putting window is controlled.
Preferably, the player further includes respond module, and the respond module is asked in response to opening or closing for user
It asks, opening or closing for AI interface is controlled.
Beneficial effects of the present invention:
1. the record data library that the present invention passes through the anonymous operation for recording user and forms user, can be according to most of
The semantic content of user or the video assists the user to play out planning or attention management to currently playing video.
2. the present invention is by the video semanteme study to knowledge point or based on a large number of users operation note, to active user
Video is played to carry out intelligent cutting or recommend associated video to user, solve user frequently looked back when learned lesson and
The problem of continually searching for associated video reduces the time loss of user, improves learning efficiency.
3. the present invention is by the video semanteme to instructional video or the machine learning based on a large number of users operation note, to working as
The video content of preceding user understands and assesses, and detect understanding of the user to current video content by investigating module,
To realize the intelligent review function of user, the learning quality of user is improved.
4. the present invention is by the video semanteme to instructional video or the machine learning based on a large number of users operation note, according to
User's attention model, the prompting in terms of the pith of video or crucial segment assist to make vision or the sense of hearing, thus
Improve the learning efficiency of user.
Detailed description of the invention
Fig. 1 is one of the structural block diagram of player in the embodiment of the present invention；
Fig. 2 is the two of the structural block diagram of the player in the embodiment of the present invention；
Fig. 3 is the playback method flow chart of the player in the embodiment of the present invention；
Fig. 4 is the flow chart that the player in the embodiment of the present invention is interacted with AI；
Fig. 5 is the flow chart that the record data library in the embodiment of the present invention is interacted with AI；
Fig. 6 is the work flow diagram of the knowledge point presentation module in the embodiment of the present invention；
Fig. 7 is the work flow diagram of the video knowledge point module in the embodiment of the present invention；
Fig. 8 is the work flow diagram that the knowledge point in the embodiment of the present invention divides automatically；
Fig. 9 is the work flow diagram that user's attention in the embodiment of the present invention is reminded；
Figure 10 is the work flow diagram that the knowledge point in the embodiment of the present invention is recommended automatically；
Figure 11 is the work flow diagram that the knowledge point in the embodiment of the present invention is investigated；
Figure 12 a- Figure 12 e is the interface layout illustrated example in the embodiment of the present invention.
Specific embodiment
The technical solution proposed in order to better understand the present invention, with reference to the accompanying drawing with specific embodiment to this hair
It is bright to be further elaborated.
A kind of playback method including recording the play operation of user, and constructs the operation note of each user；According to operation
Record assists the Automobile driving of the personalized planning and/or optimization user of user's play operation during video playing.
In an embodiment of the present invention, the play operation includes: the exhibition of the layout of broadcast window, broadcast window content
Existing, the control of playback progress, the control of broadcasting speed.
In an embodiment of the present invention, the layout of the broadcast window includes quantity, size, position, the shape of broadcast window
The control of shape.
In an embodiment of the present invention, the control of the playback progress includes whole story position, the play time that window plays
Control.
In an embodiment of the present invention, the broadcast window content show including automatically generate and/or show with currently
The segment of the relevant video of broadcasting content, or with the associated local video of the video or network video.
In an embodiment of the present invention, the foundation of the personalized planning for assisting user's play operation includes at least the use
One of the history play operation at family, the play operation of same video other users, video semanteme of the video.Further
, it is described assist user's play operation planning foundation priority be followed successively by from high to low user history play operation,
The semantic analysis of the play operation of other users, the video in same video.
In an embodiment of the present invention, Automobile driving of the optimization user during video playing includes: that basis is worked as
The content of preceding video clips and the segment carry out the prompting or triggering use of different modes in the significance level of entire video to user
The planning of family play operation.
As shown in Figure 1, a kind of player, comprising:
Playing module: encoding and decoding are carried out to the video file that plays of needs, file format conversion, color format, image are pressed
Scaling adjustment；
Control module: for the business function of control system, being broadly divided into two kinds of Control Coolings, artificial and automatic control.
Manual control includes: pause, broadcasting, intelligence AI switch, cutting video clips, maximization, minimum etc.；System automatically controls packet
Include: automatic caching, knowledge point are recommended automatically, special efficacy reminds (audio, vision), pop-up examination question etc..
Video cache module: caching currently playing video, when 1-3 minutes a length of, the storage of caching of caching
Position is local.The purpose of caching is the fluency (avoiding Caton phenomenon) for improving user and watching video.
Module is presented in knowledge point: video clips (knowledge point) corresponding with currently playing video are presented, it can be specified to playing
Video clips (knowledge point), delete specified video clips (knowledge point), can be realized by click " cutting " button newly-increased
Window is presented to knowledge point in video clips (knowledge point).
The play operation of user (learner): being recorded in local record data library by record data library, should
The information of database is saved in AI database after being sent to AI server.Play operation information includes: the length that video content divides
Degree, starting frame position, end frame position, the number that video divides, the number of viewing, pause duration etc..
Video knowledge point operation note cache module: the record data of operative knowledge point in two databases is got
And be cached to memory: first database is local operation database of record, second database be by call AI interface from
The play operation information of the same knowledge point in AI database is obtained in AI server.
AI interface: the module is the interface interacted between attention player and AI server.According to instruction type
Corresponding interface is called (when instruction type is A, knowledge point to be called to divide interface automatically；When instruction type is B, calling is known
Know the automatic recommendation interface of point；When instruction type is C, user's attention is called to remind interface；When instruction type is D, call
Investigate interface in knowledge point).
AI server-side (server): notice that power module, knowledge point are recommended automatically including the automatic division module in knowledge point, user
Module is investigated in module, knowledge point；Wherein, an instructional video the automatic division module in knowledge point: is divided into several views automatically
Frequency segment (knowledge point)；The foundation of division is to be divided video manually excessively to the instructional video according to n user (learner)
Low-level image feature in segment (knowledge point) between successive frame.
User's attention reminding module: user (learner) AI real-time detection attention during playing instructional video
Feature (visual attention feature and auditory attention feature), once detect that player can make corresponding audio or vision mentions
It wakes up.Audio is reminded: issuing the sound reminded or pop-up when such as important content or difficult point content are arrived in user (learner) study
Animation reminds user.
The automatic recommending module in knowledge point: whether user (learner) AI real-time detection during playing instructional video deposits
In the similar knowledge point that needs are recommended, once detect that player can recommend similar knowledge point.
Investigate module in knowledge point: when user (learner) has learnt a knowledge point, AI detects the knowledge point automatically
Examination paper.The examination paper of the player meeting automatic spring knowledge point can continue under study if user (learner) answers correctly
One knowledge point.If user (learner) erroneous answers can allow user (learner) to make two kinds of selections, the first selection weight
Newly learn the knowledge point, second of selection is to answer a question again.
As shown in Fig. 2, in some embodiments of the invention, user response part is separated from control module,
Whether open AI for separate responses user: if opening AI, video knowledge point cache module and control module can be with the friendships of AI
Mutually；Otherwise, it disconnects video knowledge point cache module and control module can be with the interaction of AI.
As shown in figure 3, in an embodiment of the present invention, user is as described below using the overall flow of player:
It is determined whether to enable AI: when user's selection video file, judging whether user opens AI mode: if having turned on AI
Mode then gets the record data of operative knowledge point from AI interface module and is cached to memory；If not opening AI mould
Formula, the then record data that operative knowledge point is got from record data library are cached to memory；
It obtains the operation note and presentation in memory: selecting the mark (ID) of video file to obtain from caching according to user
To corresponding operation note, the start frame and end frame of each segment after the cutting of the video is contained in operation note, so
Afterwards according to these data cutting video clips (knowledge point), the video clips (knowledge point) after cutting are put into knowledge point and are presented
Window；
The autonomous cutting of user: it is pressed if user (learner) clicks " cutting " on graphical interface of user (broadcast window)
The operation of button then carries out cutting video clips (knowledge point to video according to the starting frame position of user's selection and end frame position；
Then the video clips segmented (knowledge point) are shown in knowledge point and window is presented；
Knowledge point is presented: it is presented in window in knowledge point, if user (learner) carries out delete operation to certain knowledge point,
Then the knowledge point will delete in the window and can also delete the record from record data library；If user's (study
Person) operation is played out to certain knowledge point, then the knowledge point can be played out in broadcast window；Knowledge point position in the window
It can be with free surface jet, i.e., by the size of user's self-setting window, position, shape；
The training of AI model: AI server end gets training data from AI database first, and AI database includes view
Frequently, operation note of the user to the video；Then using the data as the input of AI module, (the automatic division module in knowledge point is known
Know the automatic recommending module of point, user's attention reminding module, knowledge point investigation), it is then trained, is obtained after training corresponding
(the automatic partitioning model in knowledge point, the automatic recommended models in knowledge point, user's attention remind model, knowledge point to investigate mould to model
Type)；Model after training can be improved and be optimized by the expansion and optimization of AI database.
Server responds service request: sending AI interface, AI interface in real time for the instruction of user and current video segment
Corresponding interface is called according to the type of instruction, then send a request to AI server and AI server to be received is waited to return to number
According to；If instruction type is A, knowledge point is called to divide interface automatically, this interface can transmit a request to AI server, AI service
Device calls the automatic partitioning model in knowledge point and result returns to AI interface；Otherwise, if instruction type is B, knowledge point is called
Automatic to recommend interface, this interface can transmit a request to AI server, the automatic recommended models in AI server calls knowledge point and result
Return to AI interface；Otherwise, if instruction type is C, user's attention is called to remind interface, this interface can be transmit a request to
AI server, AI server calls user's attention reminds model and result returns to AI interface；Otherwise, if instruction type is
D then calls knowledge point to investigate interface, this interface can transmit a request to AI server, and model is investigated in AI server calls knowledge point
And result returns to AI interface.
Play buffering: judge automatically whether meet caching condition according to network speed situation: if meeting, will be cached in video
Then local disk is played out from local disk；If do not met, video is directly played out.
As shown in figure 4, player and the detailed process that AI server interacts are as described below:
Select video file: user selects video file first, then judges whether user opens AI mode, if do not opened
AI mode is opened, then will not be interacted with AI；
It is determined whether to enable AI: if having turned on AI model, needing further to judge broadcast window with the presence or absence of broadcasting
Video；It is interacted if it does not exist, then will not trigger with AI, carries out normal play control；
AI service starting: if there is played video, it sends AI in real time by the instruction of user and current video segment
Interface, AI interface call corresponding interface according to the type of instruction, then send a request to AI server and AI to be received is waited to take
Business device returned data；If instruction type is A, association knowledge point interface is called, this interface can transmit a request to AI server,
Simultaneously result returns to AI interface to AI server calls association knowledge point model；Otherwise, if instruction type is B, knowledge is called
Point is automatic to divide interface, this interface can transmit a request to AI server, and the automatic partitioning model in AI server calls knowledge point is simultaneously tied
Fruit returns to AI interface；Otherwise, if instruction type is C, user's attention is called to remind interface, this interface can send request
To AI server, AI server calls association user attention reminds model and result returns to AI interface；Otherwise, if instruction
Type is D, then knowledge point is called to investigate interface, this interface can transmit a request to AI server, and AI server calls knowledge point is examined
It examines model and result returns to AI interface.
As shown in figure 5, in record data library and AI database, the record data that user generates has included:
Beginning frame position, end frame position, the length of video clips (knowledge point), the number of playing video file, broadcasting video clips (are known
Know point) number, suspending count, pause duration, delete video clips (knowledge point) etc..Video clips (knowledge point) refer to some
The unique corresponding instructional video in knowledge point.Operation note needs acute data cleansing before AI database.
Data cleansing: data cleansing is carried out to the data after acquisition, mainly for deviation truth or invalid data
It is cleaned.Deviate the citing 1 of the data information of truth: pause duration is too long (more than 10 minutes)；Citing 2: when broadcasting
Long too short (only playing 3 seconds)；Citing 3: the length of video clips (knowledge point) is too short (only 3 seconds)；Citing 4: it broadcasts
Device host process is killed the user operation records data for causing not being available；
As shown in fig. 6, the detailed process that knowledge point is presented is as follows:
Select video and it is determined whether to enable AI: user selects video file first, then judges that user (learner) is
No unlatching AI mode gets all starting frame positions from caching and terminates frame position, root if having turned on AI mode
Cutting video clips (knowledge point) operation, the video clips that will have been segmented are carried out to video file according to accessed frame position
(knowledge point) is added to knowledge point and window is presented and record data library is recorded in operation behavior；
Slicing operation: when not opening AI mode, if user (learner) has carried out cutting video clips (knowledge point)
Operation can then obtain starting frame position and terminate frame position；If the starting frame position got, which is greater than, terminates frame position, no
Allow to carry out slicing operation (slicing operation failure)；If the starting frame position got, which is less than, terminates frame position, cut
Divide video clips (knowledge point) operation, the video clips segmented (knowledge point) are added to knowledge point presentation window and incite somebody to action
Record data library is recorded in operation behavior；
Knowledge point deletion: delete operation is triggered after choosing some knowledge point, it will delete from record data library
The record information；
Knowledge point plays: choosing play operation of setting out behind some knowledge point, it will played out in broadcast window and
Record data records the behavior in library；
As shown in fig. 7, in an embodiment of the present invention, the detailed process of knowledge point operation note caching is as follows:
It is determined whether to enable AI: first determining whether user (learner) opens AI mode；
AI data buffer storage: if having turned on AI mode, by adjusting AI interface to get operative knowledge from AI database
The record data of point is cached to memory；
Local data banked cache: if not opening AI mode, operation is got from local record data library and is known
The record data for knowing point is cached to memory.
As shown in figure 8, in an embodiment of the present invention, the detailed process that knowledge point divides automatically is as follows:
Data prediction: it is m corresponding that the corresponding n user of each video file is got from AI database first
The starting frame position and end frame location information of video clips (knowledge point), then constantly get n by iterative process
The numerical value of the starting frame position of corresponding i-th of the video clips (knowledge point) of user and the numerical information for terminating frame position, then divide
N starting point/end frame positional value mode and the numerical information Ji Suan not be saved；
Training network: the starting point preserved according to the first step/end frame position numerical information to video file into
Row is cut into m video clips (knowledge point)；Using each frame of the video clips (knowledge point) as convolutional neural networks CNN's
Input, learns the low-level image feature of each frame by convolution kernel；Video clips (knowledge point) is extracted according to the low-level image feature of each frame
The feature vector of continuous frame sequence and by this feature vector starting point/end frame position corresponding with video clips (knowledge point)
Numerical value establish one-to-one relationship, be finally saved into the automatic partitioning model in knowledge point；It is more preferable in order to allow network to do, it needs
Further to adjust network parameter (learning rate, weight, the network number of plies, the number of neuron, the number of iterations, each round small lot
The size etc. of data) to optimize training pattern；
Respond request: user selects video file first, then judges whether user opens AI mode, if not opening AI
Mode cannot then be interacted with AI；If having turned on AI mode, when AI interface to instruction type be A when call know
Know the automatic division interface of point；Video file ID is sent to AI server by the interface, and then AI server will be according to video file
ID finds corresponding video file from AI database and is transmitted to the automatic partitioning model in knowledge point；The model can export multiple groups view
The start frame of frequency segment (knowledge point)/end frame location information gives AI server, and AI server returns result to AI interface mould
The knowledge point of block divides interface automatically；
Return the result: knowledge point divides interface to returning the result automatically, and judgement returns the result whether have starting point/knot
The frame location information of beam spot；If so, then carrying out cutting video clips according to the numerical information of starting point/end point frame position
(knowledge point) and it will be shown in knowledge points, and window is presented；If nothing, terminate this process；
As shown in figure 9, in an embodiment of the present invention, the process that user's attention is reminded is as described below:
Step 1: first from the starting frame position got in AI database in user (learner) record data
Video is carried out the data for being cut into video clips (knowledge point) by numerical value and the data for terminating frame position accordingly；
Step 2: video clips (knowledge point) to be carried out to the separation of audio and video (frame)；
Step 3: inputting frame as visual attention (NVT) algorithm, corresponding vision is obtained after the algorithm process
Attention notable figure；
Step 4: being inputted audio as auditory attention (Kayser) algorithm, obtained after the algorithm process corresponding
Auditory attention notable figure；
Step 5: getting the data information of special-effect information table from AI database；
Machine is merged step 6: establishing between visual attention notable figure, auditory attention notable figure and special-effect information table
System；
Step 7: training result, which is stored in user's attention, reminds model；
Step 8: user selects video file first, then judge whether user opens AI mode, if not opening AI mould
Formula cannot then be interacted with AI；
Step 9: teaching video contents segment is sent in real time, when the finger that AI interface arrives if having turned on AI mode
Type is enabled to call user's attention to remind interface when C；
Step 10: user's attention reminds interface that video clips are sent to AI server, then AI server is by video
Segment is transmitted to user's attention and reminds model；The model exports corresponding audio/visual special efficacy and gives AI server as a result,
AI server returns result to user's attention and reminds interface；
Step 11: user's attention after reminding interface to receive result judges whether that audio is needed to remind, if it is desired, then
Carry out the prompting of sound special efficacy；If it is not required, then further determining whether to need visual alerts；If it is required, then carrying out vision
Special efficacy is reminded；Instructional video segment is obtained if you do not need to then returning from new；
As shown in Figure 10, the detailed process that knowledge point is recommended automatically is as follows:
Step 1: first from the starting frame position got in AI database in user (learner) record data
Numerical value and the data for terminating frame position accordingly；
Step 2: video to be carried out to the data for being cut into video clips (knowledge point) according to the result of the first step；
Step 3: obtaining a video clips (knowledge point), then further drawn according to the frequency of voice and mute periods
It is divided into multiple video clips；
Step 4: obtaining a video clips in third step；
Step 5: the video clips that third step is got carry out the lock out operation of audio and video；
Step 6: audio data is changed into text by Iflytek speech-to-text interface；
Step 7: video (frame sequence) is got corresponding verbal description by NICv2 model；
Step 8: the result of the 6th step and the 7th step is segmented Chinese by ICTCLAS technology；
Step 9: the result of the 8th step is got term vector feature by Word2vec model, and again from third step
Start next iteration process；
Step 10: establishing the corresponding multiple term vector feature phases of video clips (knowledge point) after second step iteration
The incidence relation answered, and the next iteration process since second step again；
Step 11: according to the tenth step as a result, establish the matching mechanisms an of term vector Yu video clips (knowledge point),
The purpose of this mechanism is according to term vector characteristic matching to corresponding video clips (knowledge point)；
Step 12: training result is saved in the automatic recommended models in knowledge point；
Step 13: judging whether user opens AI mode, if not opening AI mould when user selects video file
Formula, then cannot interact with AI, directly control broadcasting by control module；
Step 14: instructional video segment is sent in real time, when the instruction that AI interface arrives if having turned on AI mode
Knowledge point is called to recommend interface automatically when type is B；
Step 15: knowledge point recommends interface that video content segment is sent to AI server automatically, then AI server
Video clips are transmitted to the automatic recommended models in knowledge point；The model exports corresponding examination question and gives AI server, AI as a result
Server returns result to knowledge point and recommends interface automatically；
Step 16: the knowledge for judge whether there is recommendation after interface to result is recommended in knowledge point automatically
Point, if it is present recommending similar knowledge point；Instructional video segment is obtained if it does not exist, then returning from new.
As shown in figure 11, in an embodiment of the present invention, it is as described below to investigate process for knowledge point:
Step 1: first from the starting frame position got in AI database in user (learner) record data
Numerical value and the data for terminating frame position accordingly；
Step 2: video to be carried out to the data for being cut into video clips (knowledge point) according to the result of the first step；
Step 3: iteration every first once obtains a video clips (knowledge point), then according to the frequency of voice and quiet
The sound period is further divided into multiple video clips；
Step 4: iteration every first once obtains a video clips in third step；
Step 5: the video clips that third step is got carry out the lock out operation of audio and video；
Step 6: audio data is changed into text by Iflytek speech-to-text interface；
Step 7: video (frame sequence) is got corresponding verbal description by NICv2 model；
Step 8: the result of the 6th step and the 7th step is segmented Chinese by ICTCLAS technology；
Step 9: the result of the 8th step is got term vector feature by Word2vec model, and again from third step
Start next iteration process；
Step 10: establishing the corresponding multiple term vector feature phases of video clips (knowledge point) after second step iteration
The incidence relation answered, and the next iteration process since second step again；
Step 11: establishing examination paper and video according to the result of the tenth step and the data obtained from examination paper database
Matching mechanisms between segment (knowledge point), the purpose of this mechanism are according to term vector characteristic matching to corresponding examination question；
Step 12: training result is saved in the automatic recommended models in knowledge point；
Step 13: judge whether user opens AI mode when user selects video file, if not opening AI mode,
It cannot then be interacted with AI；
Step 14: sending teaching video contents segment in real time if having turned on AI mode, being arrived when AI interface
Knowledge point is called to investigate interface when instruction type is D；
Step 15: interface is investigated in knowledge point is sent to AI server for video content segment, then AI server will be regarded
Frequency content segments are transmitted to knowledge point and investigate model；The model exports corresponding examination question and gives AI server, AI service as a result
Device returns result to knowledge point and investigates interface；
Step 16: knowledge point carries out judging whether to have learnt the knowledge point after investigating interface to result, such as
Fruit is not finished, then returns from new and obtain teaching video contents segment；
Step 17: the examination question for judging whether there is Knowledge Relation is carried out if finished, if it does not exist, then
It returns from new and obtains teaching video contents segment；If existing, the examination question of association knowledge point is popped up；
Step 18: judging whether user (learner) answers correctly, if answered correctly, next knowledge point is carried out
Until terminating；If answer is incorrect, user (learner) is allowed whether to reform examination question, if selection is reformed, popped up again
The examination question；If do not reformed, the knowledge point is replayed.
Above-mentioned steps are merely illustrative, and specific step is triggered by user and determined.
As shown in attached drawing 12a, in an embodiment of the present invention, window layout can be adjusted as needed.Upside a-quadrant
Present be all video clips (knowledge point) cut out from currently playing video or by AI generate with currently broadcast
The relevant video of the video file put or video clips, referred to as " knowledge point presentation window "；What intermediate B area was presented is currently to broadcast
The video put, referred to as " broadcast window "；The rolling of currently playing progress bar and selection front and back frame position is shown in the downside region C
Dynamic item, referred to as " playback progress window "；The lower side region D be shown trigger various operations function button (such as: pause,
The function buttons such as broadcasting, cutting, advance, retrogressing, AI mode), referred to as " operation triggering window "；
In addition to this, as shown in Figure 12 b-12e, the area A, the area B, the area C, the quantity of the window in the area D, size, position, shape position
Free adjustment can be carried out to the region by user or AI module by setting.
The above is only specific embodiments of the present invention, are not limited the scope of protection of the present invention with this；Do not violating this hair
Made any replacement and improvement, category protection scope of the present invention on the basis of bright design.
Claims (15)
1. a kind of video broadcasting method characterized by comprising
The play operation of user is recorded, and constructs the operation note of each user；
The note of the personalized planning and/or optimization user of the play operation of user during video playing is assisted according to operation note
Power of anticipating distribution.
2. video broadcasting method according to claim 1, which is characterized in that the play operation includes: broadcast window
Layout, the showing of broadcast window content, the control of playback progress, the control of broadcasting speed.
3. video broadcasting method according to claim 2, which is characterized in that the layout of the broadcast window includes playing window
The control of the quantity, size, position, shape of mouth.
4. video broadcasting method according to claim 2, which is characterized in that the control of the playback progress is broadcast including window
The whole story position put, play time control.
5. video broadcasting method according to claim 2, which is characterized in that the broadcast window content shows including certainly
The dynamic segment generated and/or show video relevant to currently playing content,
Or with the associated local video of the video or network video.
6. video broadcasting method according to claim 1, which is characterized in that the individual character of the play operation for assisting user
The foundation for changing planning includes at least the history play operation of the user, the play operation of same video other users, the video
One of video semanteme.
7. video broadcasting method according to claim 6, which is characterized in that the planning of the play operation for assisting user
Foundation priority be followed successively by from high to low the history play operation of user, the play operation of other users in same video,
The semantic analysis of the video.
8. video broadcasting method according to claim 1, which is characterized in that the optimization user is during video playing
Automobile driving includes: to be carried out not in the significance level of entire video to user according to the content and the segment of current video segment
With the prompting of mode or the planning of triggering user's play operation.
9. a kind of player, including playing module, control module, video cache module, the playing module to video file into
The layout of row broadcast window, the showing of windows content, progress, broadcasting speed are controlled, which is characterized in that
The player further include record data library, AI interface, knowledge point present module, the control module respectively with broadcast
Module connection is presented in amplification module, video cache module, AI interface, knowledge point, is configured to respond to video cache module, AI connects
The request drive control module that module is presented in mouth, knowledge point controls currently playing video；
The record data library is configured as the play operation of record user；
The AI interface is configured as the interface of connection player and AI server, in response to the request of user, calls corresponding
AI service；
The knowledge point is presented module and is configured to respond to the request of user and calls control module, will be with currently playing video
The video clips of response are presented to user.
10. player according to claim 9, which is characterized in that the AI service includes according to knowledge mapping or user
Operation note current video is divided into multiple segments, according to currently playing video from trend user recommend associated video, root
User's attention is reminded according to the content of currently playing video, pushes and works as to user according to currently playing video clips
The relevant knowledge point examination paper of preceding video clips.
11. player according to claim 10, which is characterized in that the player includes that knowledge point divides mould automatically
Module is investigated in block, the automatic recommending module in knowledge point, user's attention reminding module, knowledge point,
Current video is divided into multiple according to the operation note of knowledge mapping or user by the automatic division module in knowledge point
It is disconnected；
The automatic recommending module in knowledge point recommends associated video from trend user according to currently playing video；
User's attention reminding module reminds user's attention according to the content of currently playing video；
It investigates module and is known according to currently playing video clips to user's push is relevant to current video segment in the knowledge point
Know point examination paper.
12. player according to claim 9, which is characterized in that the player further includes video cache module, described
Video cache module caches currently playing video.
13. player according to claim 9, which is characterized in that the player includes a main broadcast window, at least
One from broadcast window,
The main broadcast window is used to play the video file of user's request；
The video clips or relevant to currently playing window of the video file for being used to play user's request from broadcast window
Video file.
14. player according to claim 13, which is characterized in that the main broadcast window can pass through from broadcast window
User or AI interface module control the layout of current video broadcast window.
15. according to any player of claim 9 or 12 or 13, which is characterized in that the player further includes response
Module, the respond module open or close request in response to user's, control opening or closing for AI interface.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811054952.6A CN109068178A (en) | 2018-09-11 | 2018-09-11 | A kind of video broadcasting method and player |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811054952.6A CN109068178A (en) | 2018-09-11 | 2018-09-11 | A kind of video broadcasting method and player |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109068178A true CN109068178A (en) | 2018-12-21 |
Family
ID=64761155
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811054952.6A Pending CN109068178A (en) | 2018-09-11 | 2018-09-11 | A kind of video broadcasting method and player |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109068178A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110337022A (en) * | 2019-06-11 | 2019-10-15 | 福建天泉教育科技有限公司 | Video variable playback method based on attention rate, storage medium |
| CN110784768A (en) * | 2019-10-17 | 2020-02-11 | 珠海格力电器股份有限公司 | Multimedia resource playing method, storage medium and electronic equipment |
| CN112507792A (en) * | 2020-11-04 | 2021-03-16 | 华中师范大学 | Online video key frame positioning method, positioning system, equipment and storage medium |
| CN113051379A (en) * | 2021-02-24 | 2021-06-29 | 南京审计大学 | Knowledge point recommendation method and device, electronic equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006317548A (en) * | 2005-05-10 | 2006-11-24 | Advanced Telecommunication Research Institute International | Educational content generating device |
| CN105095504A (en) * | 2015-08-28 | 2015-11-25 | 广东小天才科技有限公司 | Learning habit based learning content recommendation method, apparatus and system |
| CN105117381A (en) * | 2015-08-28 | 2015-12-02 | 上海第九城市教育科技股份有限公司 | Method and system for generating interactive multimedia courseware |
| CN105139706A (en) * | 2015-09-11 | 2015-12-09 | 康佳集团股份有限公司 | Online education curriculum interaction method and system based on intelligent television |
| CN106534901A (en) * | 2016-11-23 | 2017-03-22 | 广东小天才科技有限公司 | Method for pushing teaching resources and user terminal |
| CN106559709A (en) * | 2016-11-23 | 2017-04-05 | 广东小天才科技有限公司 | A kind of analysis method and server of study difficult point |
-
2018
- 2018-09-11 CN CN201811054952.6A patent/CN109068178A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006317548A (en) * | 2005-05-10 | 2006-11-24 | Advanced Telecommunication Research Institute International | Educational content generating device |
| CN105095504A (en) * | 2015-08-28 | 2015-11-25 | 广东小天才科技有限公司 | Learning habit based learning content recommendation method, apparatus and system |
| CN105117381A (en) * | 2015-08-28 | 2015-12-02 | 上海第九城市教育科技股份有限公司 | Method and system for generating interactive multimedia courseware |
| CN105139706A (en) * | 2015-09-11 | 2015-12-09 | 康佳集团股份有限公司 | Online education curriculum interaction method and system based on intelligent television |
| CN106534901A (en) * | 2016-11-23 | 2017-03-22 | 广东小天才科技有限公司 | Method for pushing teaching resources and user terminal |
| CN106559709A (en) * | 2016-11-23 | 2017-04-05 | 广东小天才科技有限公司 | A kind of analysis method and server of study difficult point |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110337022A (en) * | 2019-06-11 | 2019-10-15 | 福建天泉教育科技有限公司 | Video variable playback method based on attention rate, storage medium |
| CN110337022B (en) * | 2019-06-11 | 2022-04-12 | 福建天泉教育科技有限公司 | Attention-based video variable-speed playing method and storage medium |
| CN110784768A (en) * | 2019-10-17 | 2020-02-11 | 珠海格力电器股份有限公司 | Multimedia resource playing method, storage medium and electronic equipment |
| CN112507792A (en) * | 2020-11-04 | 2021-03-16 | 华中师范大学 | Online video key frame positioning method, positioning system, equipment and storage medium |
| CN112507792B (en) * | 2020-11-04 | 2024-01-23 | 华中师范大学 | Online video key frame positioning method, positioning system, equipment and storage medium |
| CN113051379A (en) * | 2021-02-24 | 2021-06-29 | 南京审计大学 | Knowledge point recommendation method and device, electronic equipment and storage medium |
| CN113051379B (en) * | 2021-02-24 | 2023-08-04 | 南京审计大学 | Knowledge point recommendation method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109068178A (en) | A kind of video broadcasting method and player | |
| US10528623B2 (en) | Systems and methods for content curation in video based communications | |
| CN107633719B (en) | Anthropomorphic image artificial intelligence teaching system and method based on multi-language human-computer interaction | |
| CN109416816B (en) | Artificial intelligence system supporting communication | |
| US20190025906A1 (en) | Systems and methods for virtual reality-based assessment | |
| CN110364146B (en) | Speech recognition method, speech recognition device, speech recognition apparatus, and storage medium | |
| CN106649704A (en) | Intelligent dialogue control method and intelligent dialogue control system | |
| CN111414462B (en) | Dialogue statement determination method, device, computer equipment and medium | |
| CN109996091A (en) | Generate method, apparatus, electronic equipment and the computer readable storage medium of video cover | |
| US11883730B2 (en) | Systems, devices, and methods employing the same for enhancing audience engagement in a competition or performance | |
| CA2504984A1 (en) | Survey system | |
| CN104572852B (en) | The recommended method and device of resource | |
| CN108470485A (en) | Scene-type Training Methodology, device, computer equipment and storage medium | |
| CN110069707A (en) | A kind of artificial intelligence self-adaption interactive tutoring system | |
| Hulten | Building Intelligent Systems: A Guide to Machine Learning Engineering | |
| DE202017104849U1 (en) | Systems and media for presenting a user interface custom for a predicted user activity | |
| CN110812843A (en) | Interaction method and device based on virtual image and computer storage medium | |
| JP2007334732A (en) | Network system and network information transmission/reception method | |
| CN109829064A (en) | Media resource shares and playing method and device, storage medium and electronic device | |
| CN113610680A (en) | AI-based interactive reading material personalized recommendation method and system | |
| CN112131361B (en) | Answer content pushing method and device | |
| CN107357416A (en) | A kind of human-computer interaction device and exchange method | |
| Hulten | Building Intelligent Systems | |
| CN108334518A (en) | A kind of advertisement loading method and device | |
| US20200258047A1 (en) | System and method for online sales competition and game |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181221 |