CN109344287A - A kind of information recommendation method and relevant device - Google Patents
A kind of information recommendation method and relevant device Download PDFInfo
- Publication number
- CN109344287A CN109344287A CN201811032777.0A CN201811032777A CN109344287A CN 109344287 A CN109344287 A CN 109344287A CN 201811032777 A CN201811032777 A CN 201811032777A CN 109344287 A CN109344287 A CN 109344287A
- Authority
- CN
- China
- Prior art keywords
- image
- moving image
- user
- audio
- frequency content
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/23—Recognition of whole body movements, e.g. for sport training
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2431—Multiple classes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
- G06V20/42—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items of sport video content
Abstract
The embodiment of the present application discloses a kind of information recommendation method and relevant device, comprising: obtains the moving image of user；The moving image is input in the image classification model trained, determines the class probability of the moving image；According to the class probability, the motor pattern of the user is determined；The motor pattern is sent to data server, the motor pattern is used to indicate the data server and searches the audio-video frequency content to match；Receive the audio-video frequency content that the data server is sent；The audio-video frequency content is played out.Using the embodiment of the present application, the convenience of operation can be improved.
Description
Technical field
This application involves electronic technology field more particularly to a kind of information recommendation methods and relevant device.
Background technique
More and more users like body-building, participate in various movements, user can be to listen to music in movement, or can see
It sees moving video content, after determining the essential of exercise in video content, the broadcasting of video content is followed to be moved.But
It is, for the operation of these music or video content, to require user's manual operation, search relevant audio-video frequency content, then
Play out, and user be for the operation of audio-video during the motion it is highly inconvenient, influence movement effects.
Summary of the invention
The embodiment of the present application provides a kind of information recommendation method and relevant device.By identification motor pattern search and
The relevant audio-video frequency content of motor pattern improves the convenient of operation so as to automatic playing audio-video content during the motion
Property.
On the one hand, the embodiment of the present application provides a kind of information recommendation method, comprising:
Obtain the moving image of user；
The moving image is input in the image classification model trained, determines that the classification of the moving image is general
Rate；
According to the class probability, the motor pattern of the user is determined；
The motor pattern is sent to data server, the motor pattern is used to indicate the data server and searches phase
Matched audio-video frequency content；
Receive the audio-video frequency content that the data server is sent；
The audio-video frequency content is played out.
Wherein, before the moving image for obtaining user, further includes:
It obtains multiple and waits for training image；
Extract it is described multiple wait for every body dynamics information to training image in training image, and determine described every to
The motor behavior classification of training image；
Described every is input to the body dynamics information of training image and the motor behavior classification wait instruct
Practice disaggregated model to be trained to obtain described image disaggregated model.
Wherein, the moving image for obtaining user includes:
Solicited message is sent to image capture device, the solicited message is used to indicate described image acquisition equipment and adopts in real time
Collect the moving image of the user；
Receive the moving image that described image acquisition equipment is sent.
Wherein, described according to the moving image, determine that the motor pattern of the user includes:
Whether the clarity for determining the moving image is more than preset threshold；
When the clarity of the moving image is more than the preset threshold, according to the moving image, institute is determined
State the motor pattern of user.
Wherein, described play out to the audio-video frequency content includes:
Obtain the voice messaging of the user；
The voice messaging is identified, determines control instruction included in the voice messaging；
According to the control instruction, the audio-video frequency content is played out.
On the other hand, the embodiment of the present application provides a kind of information recommending apparatus, comprising:
Module is obtained, for obtaining the moving image of user；
Processing module determines the movement for the moving image to be input in the image classification model trained
The class probability of image；According to the class probability, the motor pattern of the user is determined；
Sending module, for sending the motor pattern to data server, the motor pattern is used to indicate the number
The audio-video frequency content to match is searched according to server；
Receiving module, the audio-video frequency content sent for receiving the data server；
Playing module, for being played out to the audio-video frequency content.
Wherein, the processing module is also used to obtain multiple and waits for training image；Multiple wait for every in training image described in extracting
The body dynamics information to training image is opened, and determines every motor behavior classification to training image；It will be every described
It is input to and is trained to train classification models to the body dynamics information of training image and the motor behavior classification
Obtain described image disaggregated model.
Wherein, the sending module is also used to send solicited message to image capture device, and the solicited message is for referring to
Show that described image acquisition equipment acquires the moving image of the user in real time；
The receiving module is also used to receive the moving image that described image acquisition equipment is sent.
Wherein, the processing module is also used to determine whether the clarity of the moving image is more than preset threshold；Work as institute
When stating the clarity of moving image more than the preset threshold, according to the moving image, determine that the user's is described
Motor pattern.
Wherein, the playing module is also used to obtain the voice messaging of the user；The voice messaging is known
Not, control instruction included in the voice messaging is determined；According to the control instruction, the audio-video frequency content is broadcast
It puts.
Another aspect, the embodiment of the present application provide a kind of information recommendation equipment, comprising: processor, memory and communication
Bus, wherein for realizing connection communication between processor and memory, processor executes to be stored in memory communication bus
The step in a kind of information recommendation method that program provides for realizing above-mentioned first aspect.
In a possible design, information recommendation equipment provided by the present application be may include for executing in the above method
The corresponding module of behavior.Module can be software and/or be hardware.
The another aspect of the application provides a kind of computer readable storage medium, in the computer readable storage medium
It is stored with a plurality of instruction, described instruction is suitable for being loaded as processor and executing method described in above-mentioned various aspects.
The another aspect of the application provides a kind of computer program product comprising instruction, when it runs on computers
When, so that computer executes method described in above-mentioned various aspects.
Implement the embodiment of the present application, obtains the moving image of user first, moving image is input to the image trained
In disaggregated model, the class probability of moving image is determined；And according to the class probability, the movement mould of the user is determined
Formula；Secondly the motor pattern is sent to data server, the motor pattern is used to indicate the data server and searches phase
Matched audio-video frequency content；Then the audio-video frequency content that data server is sent is received；Finally to the audio-video frequency content
It plays out.Audio-video frequency content relevant to motor pattern is searched by identification motor pattern, so as to during the motion
Automatic playing audio-video content, improves the convenience of operation.
Detailed description of the invention
Technical solution in ord to more clearly illustrate embodiments of the present application, below will be to required use in embodiment description
Attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is some embodiments of the present application, for this field
For those of ordinary skill, without creative efforts, it is also possible to obtain other drawings based on these drawings.
Fig. 1 is a kind of structural schematic diagram of information recommendation system provided by the embodiments of the present application；
Fig. 2 is a kind of flow diagram for information recommendation method that the embodiment of the present application proposes；
Fig. 3 is a kind of schematic diagram of CNN disaggregated model provided in an embodiment of the present invention；
Fig. 4 is a kind of timing diagram for information recommendation method that the embodiment of the present application proposes；
Fig. 5 is a kind of structural schematic diagram of information recommending apparatus provided by the embodiments of the present application；
Fig. 6 is a kind of structural schematic diagram of information recommendation equipment provided by the embodiments of the present application.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present application, technical solutions in the embodiments of the present application carries out clear, complete
Site preparation description, it is clear that described embodiment is some embodiments of the present application, instead of all the embodiments.Based on this Shen
Please in embodiment, every other implementation obtained by those of ordinary skill in the art without making creative efforts
Example, shall fall in the protection scope of this application.
Referring to FIG. 1, Fig. 1 is a kind of structural schematic diagram of information recommendation system provided by the embodiments of the present application.The application
Information recommendation information in embodiment includes image capture device 101, user equipment 102 and data server 103, wherein
Information collecting device can be Kinect sensor, camera etc..User equipment 102 can refer to the voice for providing and arriving user
And/or the equipment of data connection, it also may be connected to the calculating equipment of laptop computer or desktop computer etc., or
Its autonomous device that can be such as personal digital assistant (Personal Digital Assistant, PDA) etc. of person, such as hand
Mechanical, electrical view machine etc..Data server 103 can be the server for being capable of providing server audio-video frequency content.Wherein, image is adopted
Collection equipment 101 is used to acquire the moving image of user；User equipment 102 sends motor pattern for determining motor pattern
Audio-video frequency content is played out after returning to audio-video frequency content to data server 103；Data server 103 is used for root
Matched audio-video frequency content is provided according to motor pattern.
Referring to FIG. 2, Fig. 2 is a kind of flow diagram for information recommendation method that the embodiment of the present application proposes.As schemed
Show, the step in the embodiment of the present application includes:
S201 obtains the moving image of user.
In the specific implementation, can receive the sign on of user's input, or receive the language for the initiation command that user issues
Message breath, when user equipment detects initiation command, user equipment can be established first with image capture device and is connected to the network,
Such as WIFI connection, solicited message then is sent to image capture device, after image capture device receives solicited message, is adjusted
Whole camera angle carries out real-time tracking monitoring to user, acquisition user moving image during exercise, and by collected movement
Image is sent to user equipment, the moving image that user equipment real-time reception image capture device is sent.
Further, after user equipment receives moving image, can determine the moving image clarity whether
More than preset threshold；When the clarity of the moving image is not above the preset threshold, user equipment can be shown
Show the prompt information of Image Acquisition failure, the reason of prompt information can be used for reminding user to search the failure of Image Acquisition.?
In this case, user equipment can send solicited message to image capture device again, so that image capture device is adopted again
Collect the moving image of user during exercise, and returns to user equipment.When the clarity of the moving image is more than described
When preset threshold, operations described below step is executed.
It should be noted that user equipment can be adopted directly for the user equipment for having image collecting function
Collect the moving image of user.
S202 determines the motor pattern of the user according to the moving image.Wherein, motor pattern may include running
Step mode, cycling mode, footrace mode etc..
In the specific implementation, multiple moving images can be input in the image classification model trained, institute is determined
State the class probability of moving image；According to the class probability, the motor pattern of the user is determined.As shown in figure 3, figure
3 be a kind of schematic diagram of CNN disaggregated model provided in an embodiment of the present invention.Image classification model can be convolutional neural networks
(Convolutional Neural Networks, CNN) disaggregated model.The CNN disaggregated model includes input layer, convolutional layer, pond
Change layer, full articulamentum and output layer.Wherein, the combination of convolutional layer and pond layer can occur repeatedly, occurring in Fig. 3 in hidden layer
Twice.It include that trained model parameter, the model parameter include each in image classification model extraction CNN disaggregated model
The bias vector of the convolution kernel of convolutional layer, the weight matrix of the bias matrix of each convolutional layer and full articulamentum and full articulamentum
Etc..
Further, it in image classification model, has pre-established between body dynamics information and action behavior classification
Corresponding relationship.After getting multiple moving images, multiple moving images can be input to image classification model, then existed
On each convolutional layer, using the convolution kernel and bias matrix of each convolutional layer to each pending area progress convolution operation and most
Body dynamics information of the moving image on each convolutional layer is extracted in great Chiization operation.Then, using the weight square of full articulamentum
Battle array and bias vector handle every moving image, obtain class probability.For example, when by big in multiple moving images
When componental movement image recognition obtains user and is in the class probabilities of running modes and is all larger than preset threshold (such as 50%), it is determined that
The user is in running modes.For another example, it is in when identifying to obtain user by most of moving image in multiple moving images
The class probability of cycling mode is all larger than preset threshold (such as 50%), it is determined that the user is in cycling mode.
In addition, can initially set up for image classification before the moving image for obtaining user to training classification mould
Type can be convolutional neural networks to train classification models.Then multiple are obtained and waits for training image, multiple is extracted and waits for training image
In every body dynamics information to training image, and determine every motor behavior classification to training image, the motor behavior
Classification can be the tag along sort of running modes, the tag along sort of cycling mode or tag along sort of footrace mode etc..Most
Described every is input to the body dynamics information of training image and the motor behavior classification afterwards and is classified to training
Model is trained to obtain image classification model.In the image classification model trained, establishes body dynamics information and move
Make the corresponding relationship between behavior classification.
In the embodiment of the present application, need to collect it is sufficiently large to training image, in order to provide trained accuracy.So
After will be input to training image in convolutional Neural net, come training convolutional nerve net by way of deep learning multilayer convolution,
After training is completed, the training data of convolutional Neural net is saved to the local of user equipment, such user equipment can be at this
Ground identifies the motor pattern of user, carries out processing identification without moving image is sent to data server, thus
The data volume for reducing information exchange, avoids network transmission from blocking.
S203 sends the motor pattern to data server, and the motor pattern is used to indicate the data server
Search the audio-video frequency content to match.
In the specific implementation, data server can pre-establish the database of motor pattern and audio-video frequency content, in data
After server receives motor pattern, the audio-video frequency content to match with motor pattern can be searched from database, then
The audio-video frequency content found is sent to user equipment.
Optionally, the use habit of the data server also available user, for example, the music type often played,
The video or the preceding video once played often played.After data server receives motor pattern, in conjunction with the user
Motor pattern and use habit, push relevant audio-video frequency content to user equipment.
S204 receives the audio-video frequency content that the data server is sent.
S205 plays out the audio-video frequency content.
In the specific implementation, dynamic music can be played when user is in running modes；When user is in cycling mode
When, video content can be played, so that user watches video content while cycling.
In the embodiment of the present application, the moving image of user is obtained first；According to the moving image, the user is determined
Motor pattern；Then the motor pattern is sent to data server, the motor pattern is used to indicate the data service
Device searches the audio-video frequency content to match；Receive the audio-video frequency content that the data server is sent；Finally to the sound
Video content plays out.Audio-video frequency content relevant to motor pattern is searched by identification motor pattern, to transport
Automatic playing audio-video content, improves the convenience of operation during dynamic.
Referring to FIG. 4, Fig. 4 is the timing diagram for another information recommendation method that the embodiment of the present application proposes.As shown,
Step in the embodiment of the present application includes:
S401, user equipment send solicited message to image capture device.
In the specific implementation, user equipment can receive the sign on of user's input, or receive the beginning that user issues
The voice messaging of order, when user equipment detects initiation command, user equipment can be established with image capture device first
Then network connection, such as WIFI connection send solicited message to image capture device.
S402, image capture device send the moving image of user to user equipment.
In the specific implementation, adjusting camera angle after image capture device receives solicited message, user is carried out real-time
Tracing and monitoring, the moving image of acquisition user during exercise, and collected moving image is sent to user equipment, Yong Hushe
The moving image that standby real-time reception image capture device is sent.
Further, after user equipment receives moving image, can determine the moving image clarity whether
More than preset threshold；When the clarity of the moving image is not above the preset threshold, user equipment can be shown
Show the prompt information of Image Acquisition failure, the reason of prompt information can be used for reminding user to search the failure of Image Acquisition.?
In this case, user equipment can send solicited message to image capture device again, so that image capture device is adopted again
Collect the moving image of user during exercise, and returns to user equipment.When the clarity of the moving image is more than described
When preset threshold, operations described below step is executed.
S404, user equipment determine the motor pattern of the user according to the moving image.
In the specific implementation, multiple moving images can be input in the image classification model trained, institute is determined
State the class probability of moving image；According to the class probability, the motor pattern of the user is determined.It include: in image
In disaggregated model, the corresponding relationship between body dynamics information and action behavior classification has been pre-established.Getting multiple fortune
After motion video, multiple moving images can be input to image classification model, then on each convolutional layer, use each volume
The convolution kernel and bias matrix of lamination carry out convolution operation and maximum pondization operation to each pending area, extract moving image
Body dynamics information on each convolutional layer.Then, every is moved using the weight matrix of full articulamentum and bias vector
Image is handled, and class probability is obtained.For example, being used when being identified by most of moving image in multiple moving images
When the class probability that family is in running modes is all larger than preset threshold (such as 50%), it is determined that the user is in running modes.Again
Such as, when by most of moving image in multiple moving images identify to obtain user be in cycling mode class probability it is big
In preset threshold (such as 50%), it is determined that the user is in cycling mode.
In the embodiment of the present application, need to collect it is sufficiently large to training image, in order to provide trained accuracy.So
After will be input to training image in convolutional Neural net, come training convolutional nerve net by way of deep learning multilayer convolution,
After training is completed, the training data of convolutional Neural net is saved to the local of user equipment, such user equipment can be at this
Ground identifies the motor pattern of user, carries out processing identification without moving image is sent to data server, thus
The data volume for reducing information exchange, avoids network transmission from blocking.
S404, user equipment send the motor pattern to data server.
S405, data server search the audio-video frequency content to match according to motor pattern.
In the specific implementation, data server can pre-establish the database of motor pattern and audio-video frequency content, in data
After server receives motor pattern, the audio-video frequency content to match with motor pattern can be searched from database, then
The audio-video frequency content found is sent to user equipment.
Optionally, the use habit of the data server also available user, for example, the music type often played,
The video or the preceding video once played often played.After data server receives motor pattern, in conjunction with the user
Motor pattern and use habit, push relevant audio-video frequency content to user equipment.
S406, data server send audio-video frequency content to user equipment.
S407, user equipment obtain the voice messaging of the user.
S408 identifies the voice messaging, determines control instruction included in the voice messaging.Wherein,
Control instruction may include switching command, volume adjustment instruction, sign on, halt instruction, fast forward command etc..
S409 plays out the audio-video frequency content according to the control instruction.
In the specific implementation, dynamic music can be played when user is in running modes；When user is in cycling mode
When, video content can be played, so that user watches video content while cycling.In audio-video frequency content playing process,
By identify obtained control instruction audio-video frequency content is switched over, F.F. or stopping etc. operation.In addition, if user changes
Become motor pattern, then can reacquire the moving image of the user, is sent to data after the motor pattern of identification user
Server, so that data server recommends new audio-video frequency content to play out again.
Referring to FIG. 5, Fig. 5 is a kind of structural schematic diagram for information recommending apparatus that the embodiment of the present application proposes.As schemed
Show, the device in the embodiment of the present application includes:
Module 501 is obtained, for obtaining the moving image of user.
In the specific implementation, can receive the sign on of user's input, or receive the language for the initiation command that user issues
Message breath, when user equipment detects initiation command, user equipment can be established first with image capture device and is connected to the network,
Such as WIFI connection, solicited message then is sent to image capture device, after image capture device receives solicited message, is adjusted
Whole camera angle carries out real-time tracking monitoring to user, acquisition user moving image during exercise, and by collected movement
Image is sent to user equipment, the moving image that user equipment real-time reception image capture device is sent.
Further, after user equipment receives moving image, can determine the moving image clarity whether
More than preset threshold；When the clarity of the moving image is not above the preset threshold, user equipment can be shown
Show the prompt information of Image Acquisition failure, the reason of prompt information can be used for reminding user to search the failure of Image Acquisition.?
In this case, user equipment can send solicited message to image capture device again, so that image capture device is adopted again
Collect the moving image of user during exercise, and returns to user equipment.When the clarity of the moving image is more than described
When preset threshold, operations described below step is executed.
It should be noted that user equipment can be adopted directly for the user equipment for having image collecting function
Collect the moving image of user.
Processing module 502 determines the fortune for the moving image to be input in the image classification model trained
The class probability of motion video；According to the class probability, the motor pattern of the user is determined.
As shown in figure 3, Fig. 3 is a kind of schematic diagram of CNN disaggregated model provided in an embodiment of the present invention.Image classification model
It can be convolutional neural networks (Convolutional Neural Networks, CNN) disaggregated model.The CNN disaggregated model packet
Include input layer, convolutional layer, pond layer, full articulamentum and output layer.Wherein, the combination of convolutional layer and pond layer can be in hidden layer
Occur repeatedly, occurring twice in Fig. 3.It include trained model parameter in image classification model extraction CNN disaggregated model,
The model parameter include the convolution kernel of each convolutional layer, the bias matrix of each convolutional layer and full articulamentum weight matrix and
The bias vector etc. of full articulamentum.
Further, it in image classification model, has pre-established between body dynamics information and action behavior classification
Corresponding relationship.After getting multiple moving images, multiple moving images can be input to image classification model, then existed
On each convolutional layer, using the convolution kernel and bias matrix of each convolutional layer to each pending area progress convolution operation and most
Body dynamics information of the moving image on each convolutional layer is extracted in great Chiization operation.Then, using the weight square of full articulamentum
Battle array and bias vector handle every moving image, obtain class probability.For example, when by big in multiple moving images
When componental movement image recognition obtains user and is in the class probabilities of running modes and is all larger than preset threshold (such as 50%), it is determined that
The user is in running modes.For another example, it is in when identifying to obtain user by most of moving image in multiple moving images
The class probability of cycling mode is all larger than preset threshold (such as 50%), it is determined that the user is in cycling mode.
In addition, can initially set up for image classification before the moving image for obtaining user to training classification mould
Type can be convolutional neural networks to train classification models.Then multiple are obtained and waits for training image, multiple is extracted and waits for training image
In every body dynamics information to training image, and determine every motor behavior classification to training image, the motor behavior
Classification can be the tag along sort of running modes, the tag along sort of cycling mode or tag along sort of footrace mode etc..Most
Described every is input to the body dynamics information of training image and the motor behavior classification afterwards and is classified to training
Model is trained to obtain image classification model.In the image classification model trained, establishes body dynamics information and move
Make the corresponding relationship between behavior classification.
In the embodiment of the present application, need to collect it is sufficiently large to training image, in order to provide trained accuracy.So
After will be input to training image in convolutional Neural net, come training convolutional nerve net by way of deep learning multilayer convolution,
After training is completed, the training data of convolutional Neural net is saved to the local of user equipment, such user equipment can be at this
Ground identifies the motor pattern of user, carries out processing identification without moving image is sent to data server, thus
The data volume for reducing information exchange, avoids network transmission from blocking.
Sending module 503, for sending the motor pattern to data server, the motor pattern is used to indicate described
Data server searches the audio-video frequency content to match.
In the specific implementation, data server can pre-establish the database of motor pattern and audio-video frequency content, in data
After server receives motor pattern, the audio-video frequency content to match with motor pattern can be searched from database, then
The audio-video frequency content found is sent to user equipment.
Optionally, the use habit of the data server also available user, for example, the music type often played,
The video or the preceding video once played often played.After data server receives motor pattern, in conjunction with the user
Motor pattern and use habit, push relevant audio-video frequency content to user equipment.
Receiving module 505, the audio-video frequency content sent for receiving the data server.
Playing module 505, for being played out to the audio-video frequency content.
In the specific implementation, the voice messaging of the user can be obtained first, then the voice messaging is identified,
Determine control instruction included in the voice messaging.Wherein, control instruction may include that switching command, volume adjustment refer to
It enables, sign on, halt instruction, fast forward command etc..Finally according to the control instruction, the audio-video frequency content is broadcast
It puts.
For example, dynamic music can be played when user is in running modes；It, can when user is in cycling mode
To play video content, so that user watches video content while cycling.In audio-video frequency content playing process, pass through knowledge
The control instruction not obtained switches over audio-video frequency content, F.F. or stopping etc. operation.In addition, if user changes movement
Mode can then reacquire the moving image of the user, be sent to data server after the motor pattern of identification user,
So that data server recommends new audio-video frequency content to play out again.
In the embodiment of the present application, the moving image of user is obtained first；According to the moving image, the user is determined
Motor pattern；Then the motor pattern is sent to data server, the motor pattern is used to indicate the data service
Device searches the audio-video frequency content to match；Receive the audio-video frequency content that the data server is sent；Finally to the sound
Video content plays out.Audio-video frequency content relevant to motor pattern is searched by identification motor pattern, to transport
Automatic playing audio-video content, improves the convenience of operation during dynamic.
With continued reference to FIG. 6, Fig. 6 is a kind of structural schematic diagram for information recommendation equipment that the embodiment of the present application proposes.Such as
Shown in figure, which may include: at least one processor 601, at least one communication interface 602, at least one
Memory 603 and at least one communication bus 604.
Wherein, processor 601 can be central processor unit, general processor, digital signal processor, dedicated integrated
Circuit, field programmable gate array or other programmable logic device, transistor logic, hardware component or it is any
Combination.It, which may be implemented or executes, combines various illustrative logic blocks, module and electricity described in present disclosure
Road.The processor is also possible to realize the combination of computing function, such as combines comprising one or more microprocessors, number letter
Number processor and the combination of microprocessor etc..Communication bus 604 can be Peripheral Component Interconnect standard PCI bus or extension work
Industry normal structure eisa bus etc..The bus can be divided into address bus, data/address bus, control bus etc..For convenient for indicate,
It is only indicated with a thick line in Fig. 6, it is not intended that an only bus or a type of bus.Communication bus 604 is used for
Realize the connection communication between these components.Wherein, the communication interface 602 of equipment is used for and other nodes in the embodiment of the present application
Equipment carries out the communication of signaling or data.Memory 603 may include volatile memory, such as non-volatile dynamic random is deposited
Take memory (Nonvolatile Random Access Memory, NVRAM), phase change random access memory (Phase
Change RAM, PRAM), magnetic-resistance random access memory (Magetoresistive RAM, MRAM) etc., can also include non-
Volatile memory, for example, at least a disk memory, Electrical Erasable programmable read only memory (Electrically
Erasable Programmable Read-Only Memory, EEPROM), flush memory device, such as anti-or flash memory (NOR
Flash memory) or anti-and flash memory (NAND flash memory), semiconductor devices, such as solid state hard disk (Solid
State Disk, SSD) etc..Memory 603 optionally can also be that at least one is located remotely from the storage of aforementioned processor 601
Device.Batch processing code is stored in memory 603, and processor 601 executes the program in memory 603.
Obtain the moving image of user；
The moving image is input in the image classification model trained, determines that the classification of the moving image is general
Rate；According to the class probability, the motor pattern of the user is determined；
The motor pattern is sent to data server, the motor pattern is used to indicate the data server and searches phase
Matched audio-video frequency content；
Receive the audio-video frequency content that the data server is sent；
The audio-video frequency content is played out.
Optionally, processor 601 is also used to perform the following operations step:
It obtains multiple and waits for training image；
Extract it is described multiple wait for every body dynamics information to training image in training image, and determine described every to
The motor behavior classification of training image；
Described every is input to the body dynamics information of training image and the motor behavior classification wait instruct
Practice disaggregated model to be trained to obtain described image disaggregated model.
Optionally, processor 601 is also used to perform the following operations step:
Solicited message is sent to image capture device, the solicited message is used to indicate described image acquisition equipment and adopts in real time
Collect the moving image of the user；
Receive the moving image that described image acquisition equipment is sent.
Optionally, processor 601 is also used to perform the following operations step:
Whether the clarity for determining the moving image is more than preset threshold；
When the clarity of the moving image is more than the preset threshold, according to the moving image, institute is determined
State the motor pattern of user.
Optionally, processor 601 is also used to perform the following operations step:
Obtain the voice messaging of the user；
The voice messaging is identified, determines control instruction included in the voice messaging；
According to the control instruction, the audio-video frequency content is played out.
Further, processor can also be matched with memory and communication interface, executed and believed in above-mentioned application embodiment
Cease the operation of recommendation apparatus.
In the above-described embodiments, can come wholly or partly by software, hardware, firmware or any combination thereof real
It is existing.When implemented in software, it can entirely or partly realize in the form of a computer program product.The computer program
Product includes one or more computer instructions.When loading on computers and executing the computer program instructions, all or
It partly generates according to process or function described in the embodiment of the present application.The computer can be general purpose computer, dedicated meter
Calculation machine, computer network or other programmable devices.The computer instruction can store in computer readable storage medium
In, or from a computer readable storage medium to the transmission of another computer readable storage medium, for example, the computer
Instruction can pass through wired (such as coaxial cable, optical fiber, number from a web-site, computer, server or data center
User's line (DSL)) or wireless (such as infrared, wireless, microwave etc.) mode to another web-site, computer, server or
Data center is transmitted.The computer readable storage medium can be any usable medium that computer can access or
It is comprising data storage devices such as one or more usable mediums integrated server, data centers.The usable medium can be with
It is magnetic medium, (for example, floppy disk, hard disk, tape), optical medium (for example, DVD) or semiconductor medium (such as solid state hard disk
Solid State Disk (SSD)) etc..
Above-described specific embodiment has carried out further the purpose of the application, technical scheme and beneficial effects
It is described in detail.Within the spirit and principles of this application, any modification, equivalent replacement, improvement and so on should be included in
Within the protection scope of the application.
Claims (12)
1. a kind of information recommendation method, which is characterized in that the described method includes:
Obtain the moving image of user；
The moving image is input in the image classification model trained, determines the class probability of the moving image；
According to the class probability, the motor pattern of the user is determined；
The motor pattern is sent to data server, the motor pattern is used to indicate the data server lookup and matches
Audio-video frequency content；
Receive the audio-video frequency content that the data server is sent；
The audio-video frequency content is played out.
2. the method as described in claim 1, which is characterized in that before the moving image for obtaining user, further includes:
It obtains multiple and waits for training image；
Extract it is described multiple wait for every body dynamics information to training image in training image, and determine described every wait train
The motor behavior classification of image；
Described every is input to the body dynamics information of training image and the motor behavior classification to training point
Class model is trained to obtain described image disaggregated model.
3. the method as described in claim 1, which is characterized in that it is described obtain user moving image include:
Solicited message is sent to image capture device, the solicited message is used to indicate described image acquisition equipment and acquires institute in real time
State the moving image of user；
Receive the moving image that described image acquisition equipment is sent.
4. the method as described in claim 1, which is characterized in that it is described according to the moving image, determine the fortune of the user
Dynamic model formula includes:
Whether the clarity for determining the moving image is more than preset threshold；
When the clarity of the moving image is more than the preset threshold, according to the moving image, the use is determined
The motor pattern at family.
5. method according to any of claims 1-4, which is characterized in that described to play out packet to the audio-video frequency content
It includes:
Obtain the voice messaging of the user；
The voice messaging is identified, determines control instruction included in the voice messaging；
According to the control instruction, the audio-video frequency content is played out.
6. a kind of information recommending apparatus, which is characterized in that described device includes:
Module is obtained, for obtaining the moving image of user；
Processing module determines the moving image for the moving image to be input in the image classification model trained
Class probability determine the motor pattern of the user and according to the class probability；
Sending module, for sending the motor pattern to data server, the motor pattern is used to indicate the data clothes
Business device searches the audio-video frequency content to match；
Receiving module, the audio-video frequency content sent for receiving the data server；
Playing module, for being played out to the audio-video frequency content.
7. device as claimed in claim 6, which is characterized in that
The processing module is also used to obtain multiple and waits for training image；Extract it is described multiple wait in training image every wait train
The body dynamics information of image, and determine every motor behavior classification to training image；Described every is schemed to training
The body dynamics information of picture and the motor behavior classification be input to be trained to obtain to train classification models it is described
Image classification model.
8. device as claimed in claim 6, which is characterized in that
The sending module is also used to send solicited message to image capture device, and the solicited message is used to indicate the figure
As acquisition equipment acquires the moving image of the user in real time；
The receiving module is also used to receive the moving image that described image acquisition equipment is sent.
9. device as claimed in claim 6, which is characterized in that
The processing module is also used to determine whether the clarity of the moving image is more than preset threshold；
When the clarity of the moving image is more than the preset threshold, according to the moving image, the use is determined
The motor pattern at family.
10. device as claim in any one of claims 6-9, which is characterized in that
The playing module is also used to obtain the voice messaging of the user；The voice messaging is identified, described in determination
Control instruction included in voice messaging；According to the control instruction, the audio-video frequency content is played out.
11. a kind of information recommendation equipment characterized by comprising memory, communication bus and processor, wherein described to deposit
Reservoir is executed for calling said program code such as any one of claim 1-5 institute for storing program code, the processor
The method stated.
12. a kind of computer readable storage medium, which is characterized in that the computer-readable recording medium storage has a plurality of finger
It enables, described instruction is suitable for being loaded by processor and executing the method according to claim 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811032777.0A CN109344287A (en) | 2018-09-05 | 2018-09-05 | A kind of information recommendation method and relevant device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811032777.0A CN109344287A (en) | 2018-09-05 | 2018-09-05 | A kind of information recommendation method and relevant device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109344287A true CN109344287A (en) | 2019-02-15 |
Family
ID=65292161
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811032777.0A Pending CN109344287A (en) | 2018-09-05 | 2018-09-05 | A kind of information recommendation method and relevant device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109344287A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110300329A (en) * | 2019-06-26 | 2019-10-01 | 北京字节跳动网络技术有限公司 | Video pushing method, device and electronic equipment based on discrete features |
| CN110991561A (en) * | 2019-12-20 | 2020-04-10 | 山东大学齐鲁医院 | Method and system for identifying images of endoscope in lower digestive tract |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101158883A (en) * | 2007-10-09 | 2008-04-09 | 深圳先进技术研究院 | Virtual gym system based on computer visual sense and realize method thereof |
| CN101388233A (en) * | 2008-10-22 | 2009-03-18 | 深圳华为通信技术有限公司 | Method for playing music based on step of user and mobile terminal |
| US7558809B2 (en) * | 2006-01-06 | 2009-07-07 | Mitsubishi Electric Research Laboratories, Inc. | Task specific audio classification for identifying video highlights |
| CN101604325A (en) * | 2009-07-17 | 2009-12-16 | 北京邮电大学 | Method for classifying sports video based on key frame of main scene lens |
| CN105702270A (en) * | 2014-11-27 | 2016-06-22 | 深圳市腾讯计算机系统有限公司 | Music playing method and apparatus |
| CN106547787A (en) * | 2015-09-22 | 2017-03-29 | 腾讯科技（深圳）有限公司 | A kind of method of music recommendation, apparatus and system |
| CN107944431A (en) * | 2017-12-19 | 2018-04-20 | 陈明光 | A kind of intelligent identification Method based on motion change |
-
2018
- 2018-09-05 CN CN201811032777.0A patent/CN109344287A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7558809B2 (en) * | 2006-01-06 | 2009-07-07 | Mitsubishi Electric Research Laboratories, Inc. | Task specific audio classification for identifying video highlights |
| CN101158883A (en) * | 2007-10-09 | 2008-04-09 | 深圳先进技术研究院 | Virtual gym system based on computer visual sense and realize method thereof |
| CN101388233A (en) * | 2008-10-22 | 2009-03-18 | 深圳华为通信技术有限公司 | Method for playing music based on step of user and mobile terminal |
| CN101604325A (en) * | 2009-07-17 | 2009-12-16 | 北京邮电大学 | Method for classifying sports video based on key frame of main scene lens |
| CN105702270A (en) * | 2014-11-27 | 2016-06-22 | 深圳市腾讯计算机系统有限公司 | Music playing method and apparatus |
| CN106547787A (en) * | 2015-09-22 | 2017-03-29 | 腾讯科技（深圳）有限公司 | A kind of method of music recommendation, apparatus and system |
| CN107944431A (en) * | 2017-12-19 | 2018-04-20 | 陈明光 | A kind of intelligent identification Method based on motion change |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110300329A (en) * | 2019-06-26 | 2019-10-01 | 北京字节跳动网络技术有限公司 | Video pushing method, device and electronic equipment based on discrete features |
| CN110991561A (en) * | 2019-12-20 | 2020-04-10 | 山东大学齐鲁医院 | Method and system for identifying images of endoscope in lower digestive tract |
| CN110991561B (en) * | 2019-12-20 | 2020-12-25 | 山东大学齐鲁医院 | Method and system for identifying images of endoscope in lower digestive tract |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110012356B (en) | Video recommendation method, device and equipment and computer storage medium | |
| US10438091B2 (en) | Method and apparatus for recognizing image content | |
| WO2016074492A1 (en) | Social platform-based data mining method and device | |
| CN106339507B (en) | Streaming Media information push method and device | |
| CN108319723A (en) | A kind of picture sharing method and device, terminal, storage medium | |
| CN107977390A (en) | Paint this recognition methods, device, system and electronic equipment | |
| CN107947951A (en) | Groups of users recommends method, apparatus and storage medium and server | |
| CN109240875B (en) | Canton analysis method and system | |
| CN110209810B (en) | Similar text recognition method and device | |
| CN111107435B (en) | Video recommendation method and device | |
| CN106789543A (en) | The method and apparatus that facial expression image sends are realized in session | |
| US20230267142A1 (en) | Multimedia content publishing method and apparatus, electronic device, and storage medium | |
| CN108304373A (en) | Construction method, device, storage medium and the electronic device of semantic dictionary | |
| CN111523324B (en) | Named entity recognition model training method and device | |
| CN107368182B (en) | Gesture detection network training, gesture detection and gesture control method and device | |
| CN104915351A (en) | Picture sorting method and terminal | |
| CN109165309B (en) | Negative example training sample acquisition method and device and model training method and device | |
| WO2019062342A1 (en) | Background application cleaning method and apparatus, and storage medium and electronic device | |
| WO2021160119A1 (en) | Information display method and device | |
| WO2012145273A1 (en) | Systems and methods for automatically determining an improved view for a visual query in a mobile search | |
| CN109344287A (en) | A kind of information recommendation method and relevant device | |
| CN110446104A (en) | Method for processing video frequency, device and storage medium | |
| WO2020135054A1 (en) | Method, device and apparatus for video recommendation and storage medium | |
| CN111259257A (en) | Information display method, system, device, electronic equipment and storage medium | |
| CN111046203A (en) | Image retrieval method, image retrieval device, storage medium and electronic equipment |
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 |