WO2019114015A1 - 一种机器人的演奏控制方法及机器人 - Google Patents

一种机器人的演奏控制方法及机器人 Download PDF

Info

Publication number
WO2019114015A1
WO2019114015A1 PCT/CN2017/117681 CN2017117681W WO2019114015A1 WO 2019114015 A1 WO2019114015 A1 WO 2019114015A1 CN 2017117681 W CN2017117681 W CN 2017117681W WO 2019114015 A1 WO2019114015 A1 WO 2019114015A1
Authority
WO
WIPO (PCT)
Prior art keywords
robot
music
performance
unit
target
Prior art date
Application number
PCT/CN2017/117681
Other languages
English (en)
French (fr)
Inventor
蔡任轩
Original Assignee
广州德科投资咨询有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 广州德科投资咨询有限公司 filed Critical 广州德科投资咨询有限公司
Publication of WO2019114015A1 publication Critical patent/WO2019114015A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
    • G10H1/00—Details of electrophonic musical instruments
    • G10H1/0033—Recording/reproducing or transmission of music for electrophonic musical instruments
    • G10H1/0041—Recording/reproducing or transmission of music for electrophonic musical instruments in coded form
    • G10H1/0058—Transmission between separate instruments or between individual components of a musical system
    • G10H1/0066—Transmission between separate instruments or between individual components of a musical system using a MIDI interface
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes

Definitions

  • the invention relates to the field of artificial intelligence technology, and in particular to a performance control method for a robot and a robot.
  • robots can also be like performers. Playing music gives people a different listening experience.
  • the robot it is possible to learn a version in which the different musicians play the same piece of music by a machine learning method, thereby obtaining a performance mode of playing the piece of music, including the performance speed of the piece, the playing strength, and the like.
  • the robot can only play in a fixed performance mode when playing the same piece of music, and the robot is played in a single mode.
  • the embodiment of the invention discloses a performance control method for a robot and a robot, which can enable the robot to adopt different performance modes in different performance situations, thereby improving the user experience of the listener.
  • the robot receives a performance instruction, the performance instruction including at least a music name of the target music piece;
  • the robot acquires an original MIDI file that matches the name of the music piece
  • the robot parses the original MIDI file to obtain a performance control parameter of the target music piece
  • the robot controls the built-in camera of the robot to capture an image, and identifies an environment in which the robot is currently located according to the image;
  • the robot adjusts the performance control parameter to a target performance control parameter that matches an environment in which the robot is currently located, and controls the robot to play the target music according to the target performance control parameter.
  • the method further includes:
  • the robot detects whether a new performance instruction is received within a preset time period, and a start time of the preset time period is a time at which the target music performance ends;
  • the robot acquires at least one to-be-played music whose similarity with the target music is higher than a preset threshold
  • the robot outputs at least one song name of the to-be-played music by voice, so that the user selects a target music to be played according to the music name of at least one of the to-be-played music pieces.
  • the method further includes:
  • the robot constructs a convolutional neural network model for identifying an environment in which the robot is located;
  • the robot acquires a massive sample image and forms a training sample set according to the massive sample image
  • the robot trains the convolutional neural network model according to the training sample set to obtain a trained convolutional neural network model
  • the robot identifies an environment in which the robot is currently located according to the image, including:
  • the robot identifies the image using the trained convolutional neural network model to identify the environment in which the robot is currently located.
  • the method further includes:
  • the robot controls the pickup to record an audio signal when the robot plays the target music piece;
  • the robot determines a noise signal from the audio signal, acquires a characteristic parameter of the noise signal, and performs noise reduction processing on the audio signal according to the characteristic parameter to obtain an audio file;
  • the robot detects whether there is a mobile terminal that establishes a communication connection with the robot;
  • the robot transmits sharing information to the mobile terminal, the sharing information including the audio file.
  • the method further includes:
  • the robot acquires position coordinates corresponding to the current environment
  • the robot associates the audio file with the position coordinates to obtain an association relationship between the audio file and the position coordinates;
  • the robot shares the association relationship to the Internet terminal, so that the Internet terminal recommends the audio file to the mobile terminal of the other user when detecting the presence of the mobile terminal of the other user at the location coordinate.
  • a second aspect of an embodiment of the present invention discloses a robot, the robot comprising:
  • a receiving unit configured to receive a performance instruction, where the performance instruction includes at least a music name of the target music piece;
  • a first obtaining unit configured to acquire an original MIDI file that matches the name of the music piece
  • a parsing unit configured to parse the original MIDI file to obtain a performance control parameter of the target music piece
  • a recognition unit configured to control a built-in camera of the robot to capture an image, and identify an environment in which the robot is currently located according to the image;
  • An adjusting unit configured to adjust the performance control parameter to a target performance control parameter that matches an environment in which the robot is currently located;
  • a performance unit configured to control the robot to play the target music piece according to the target performance performance control parameter.
  • the robot further includes:
  • a first detecting unit configured to detect whether a new performance instruction is received within a preset time period after the performance unit controls the robot to play the target music piece according to the target performance control parameter, the preset time period
  • the starting moment is the time at which the target music performance ends
  • a second acquiring unit configured to acquire, when the first detecting unit detects that the new performance instruction is not received within the preset time period, acquiring at least one of a similarity with the target music is higher than a preset threshold First to play music;
  • an output unit configured to output at least one song name of the to-be-played music by voice, so that the user selects a target music to be played according to the music name of the at least one of the to-be-played music pieces.
  • the robot further includes:
  • a building unit for constructing a convolutional neural network model for identifying an environment in which the robot is located
  • Forming a unit configured to acquire a massive sample image, and form a training sample set according to the massive sample image
  • a training unit configured to train the convolutional neural network model according to the training sample set, to obtain a trained convolutional neural network model
  • the identifying unit is specifically configured to identify the image by using the trained convolutional neural network model obtained by the training unit to identify an environment in which the robot is currently located.
  • the robot further includes:
  • a recording unit configured to control, when the performance unit plays the target music piece, a sound signal when the robot plays the target music piece;
  • a processing unit configured to determine a noise signal from the audio signal, acquire a characteristic parameter of the noise signal, and perform noise reduction processing on the audio signal according to the characteristic parameter to obtain an audio file;
  • a second detecting unit configured to detect whether there is a mobile terminal that establishes a communication connection with the robot
  • a sending unit configured to: when the second detecting unit detects that there is a mobile terminal that establishes a communication connection with the robot, send the sharing information to the mobile terminal, where the sharing information includes the audio file.
  • the first acquiring unit is further configured to acquire location coordinates corresponding to the current environment
  • the robot also includes:
  • An association unit configured to associate the audio file with the location coordinates to obtain an association relationship between the audio file and the location coordinates
  • a sharing unit configured to share the association relationship to an Internet terminal, so that the Internet terminal recommends the audio file to a mobile terminal of the other user when detecting that the mobile terminal of another user exists at the location coordinate .
  • a third aspect of the embodiment of the present invention discloses another robot, including:
  • a processor coupled to the memory
  • the processor invokes the executable program code stored in the memory to perform a performance control method of the robot disclosed in the first aspect of the embodiments of the present invention.
  • a fourth aspect of the embodiments of the present invention discloses a computer readable storage medium storing a computer program, wherein the computer program causes a computer to execute a performance control method of a robot disclosed in the first aspect of the embodiments of the present invention.
  • a fifth aspect of an embodiment of the present invention discloses a computer program product that, when executed on a computer, causes the computer to perform some or all of the steps of any of the methods of the first aspect.
  • a sixth aspect of the embodiments of the present invention discloses an application publishing platform, which is used for publishing a computer program product, wherein when the computer program product runs on a computer, the computer is caused to perform any of the first aspect.
  • the embodiment of the invention has the following beneficial effects:
  • the robot receives a performance instruction
  • the performance instruction includes at least a music name of the target music, acquires and parses the original MIDI file matching the music name, obtains performance control parameters of the target music, and controls the built-in camera of the robot.
  • the image is captured, and the current environment of the robot is recognized according to the image, and then the performance control parameter is adjusted to the target performance control parameter that matches the environment in which the robot is currently located, and the robot performance target music is controlled according to the target performance control parameter.
  • the performance control parameters can be adjusted according to the surrounding environment, so that the robot can play the music in different performance modes in different performance situations, thereby improving the user experience of the listener.
  • FIG. 1 is a schematic flow chart of a method for controlling performance of a robot disclosed in an embodiment of the present invention
  • FIG. 2 is a schematic flow chart of another performance control method for a robot disclosed in an embodiment of the present invention.
  • FIG. 3 is a schematic flow chart of still another method for controlling performance of a robot according to an embodiment of the present invention.
  • FIG. 4 is a schematic structural diagram of a robot disclosed in an embodiment of the present invention.
  • FIG. 5 is a schematic structural diagram of another robot disclosed in an embodiment of the present invention.
  • FIG. 6 is a schematic structural diagram of still another robot disclosed in an embodiment of the present invention.
  • FIG. 7 is a schematic structural diagram of still another robot disclosed in an embodiment of the present invention.
  • the embodiment of the invention discloses a performance control method for a robot and a robot, which can enable the robot to adopt different performance modes in different performance situations, thereby improving the user experience of the listener. The details are described below separately.
  • FIG. 1 is a schematic flow chart of a method for controlling performance of a robot according to an embodiment of the present invention.
  • the performance control method of the robot may include the following operations:
  • the robot receives a performance instruction, where the performance instruction includes at least a music name of the target music.
  • the performance instruction may be sent by the user through a mobile terminal that is pre-bound with the robot.
  • the performance of the performance instruction may be text or voice, which is not limited in the embodiment of the present invention.
  • the user may be a parent, and the mobile terminal pre-bound with the robot may be a parent's mobile phone. Then, when the parent goes out and the child is alone at home, the parent hopes that the robot can play the piano music for the child, thereby cultivating the child's
  • the sentiment, the child's interest in music so parents can send voice messages to the robot through the mobile phone, the voice information includes the music name of the target music that the parents want the robot to play.
  • the robot After receiving the voice information sent by the parent, the robot recognizes the voice information, and plays the piano song corresponding to the music name for the child according to the recognized music name of the target music. It can be seen that, in the embodiment of the present invention, the remote control robot can play the music for the child when the parent goes out, and the real piano performance appreciation is brought to the child, thereby improving the user experience.
  • the robot acquires an original MIDI file that matches the name of the music piece.
  • MIDI Musical Instrument Digital Interface
  • MIDI files are mainly divided into two parts: header information (Header Chunk) and track information (Track Chunk).
  • the header information mainly stores basic information in MIDI files, such as: each quarter note.
  • the number of points (Tick), the number of track information blocks, the file format code, etc., are cut, and the track information is used to store a series of MIDI messages.
  • MIDI information can be used to represent a single tone.
  • the MIDI information of the track (file format code 0) or the plurality of tracks (file format code 1) is not limited in the embodiment of the present invention.
  • the robot may generate an original MIDI file matching the music name by processing the audio file matched by the music name, specifically Said that the robot can process the audio file matching the music name to obtain the time attribute description information of the audio file, the time attribute description information is a time attribute corresponding to each note in the audio file; and then according to the preset time length Performing a framing process on the audio file to obtain a spectral center of gravity of each audio frame in the audio file, and calculating a spectral mean value of the audio file according to the center of gravity of the spectrum, and normalizing the average value of the spectrum to obtain a MIDI file parameter.
  • the MIDI parameters can be used to describe the pitch information of the audio file; the robot can generate an original MIDI file matching the music name according to the time attribute corresponding to each note in the audio file and the pitch information of the audio file. It can be seen that, in the embodiment of the present invention, the MIDI file corresponding to the audio file can be calculated and generated through the audio file, thereby improving the efficiency of the audio processing, and further improving the intelligence of the audio processing.
  • the robot parses the original MIDI file to obtain performance control parameters of the target music.
  • the robot parses the original MIDI file, and the MIDI file parameter may be obtained, wherein the MIDI file parameter may include all the notes of the target music and the interval between each note of all the notes and the start of the target music.
  • the note data, the pitch (frequency), the time value (duration), and the velocity data such as velocity, the MIDI file parameter is the performance control parameter of the target music.
  • the robot parses the original MIDI file, and can also obtain the overall timbre of the target music.
  • the robot can adjust the overall tone of the target music according to any tone selected by the user, and then adapt the corresponding parameters to the adjusted overall tone.
  • the robot controls the built-in camera of the robot to take an image, and identifies an environment in which the robot is currently located according to the image.
  • the robot can control the built-in camera to capture an image.
  • the robot can control the built-in camera to capture the panoramic image, that is, the robot can preset the shooting range and set a plurality of panoramic focus positions, and capture the plurality of the above through the built-in camera.
  • a plurality of in-focus images corresponding to the panoramic in-focus position, and the panoramic images corresponding to the preset photographing range are generated from the plurality of in-focus images.
  • the robot adjusts the performance control parameter to a target performance control parameter that matches the environment in which the robot is currently located, and controls the robot to play the target music according to the target performance control parameter.
  • the robot is controlled to play the target music according to the target performance control parameter.
  • the target performance control parameter may be converted into a pulse output signal, that is, read into the target performance control parameter.
  • Note data which is the interval (offset), pitch (frequency), time value (duration), and velocity of each note in the target song and the start of the target song, and is formed based on the note data.
  • Each of the notes corresponds to a pulse signal
  • the pulse signals of different notes correspond to different signal output channels
  • the signal output channel is associated with the motor of the robot. For example, when a robot plays a piece of music through a piano, the robot can output a pulse signal in the chronological order of each note, which can drive the motor to start, that is, the robot starts playing the target piece of music.
  • control parameters can be played according to different intelligent adjustments of the surrounding environment, so that the robot can play the music in different performance modes in different performance situations, thereby improving the user experience of the listener;
  • FIG. 2 is a schematic flowchart diagram of another method for controlling performance of a robot disclosed in an embodiment of the present invention. As shown in FIG. 2, the performance control method of the robot may include the following steps:
  • the method for controlling the performance of the robot further includes the steps 201 to 205.
  • the steps 201 to 205 refer to the detailed description of the steps 101 to 105 in the first embodiment, and details are not described herein again.
  • the robot detects whether a new performance instruction is received within the preset time period, and if so, plays a new music according to the new performance instruction, and if not, proceeds to step 207.
  • the start time of the preset time period is the time when the target music performance ends.
  • the robot is pre-configured with a user information library, and the user information database pre-stores at least one legal user's legal face image and/or legal voiceprint data, and each legal user's favorite music type Information in which the legitimate face image and/or legal voiceprint material of each legitimate user is bound to the favorite music type data.
  • the current environment of the robot is at home.
  • the robot can activate the face recognition function and/or the voiceprint recognition function to obtain the user's photo and / or voiceprint, and match the user's photo and / or voiceprint with the legal face image and / or legal voiceprint data of all legitimate users in the user information database, if the user's photo and / or voiceprint and If the legal face image and/or the legal voiceprint data of any legal user in the user information database match, the robot can retrieve the favorite music type data of the matching legitimate user in the user information database, and analyze the favorite The music type data is used to push a music that matches the favorite music type data to the user.
  • the robot when the robot activates the face recognition function and/or the voiceprint recognition function and the number of users who successfully match the legitimate users in the user information base is greater than one digit, the robot can calculate the favorite music type data of the users. The degree of coincidence to select the push song. It can be seen that, in the embodiment of the present invention, after the legal user is identified through the face and/or the voiceprint, the user can provide the personalized music recommendation according to the music style data pre-stored by the legitimate user, thereby improving the user experience.
  • the robot acquires at least one to-be-played music whose similarity with the target music is higher than a preset threshold.
  • the robot outputs at least one song name of the music to be played by voice, so that the user selects a target music to be played according to the music name of the at least one music to be played.
  • the robot may preset a preset time period, and the start time of the preset time period is the end time of the end of the previous music performance. Then, after a song is played, if the robot receives a new performance instruction within the preset time period, the robot can play the new music according to the new music name included in the new performance instruction; if the robot is in the preset The new performance command is not received within the time period, and the robot can push the to-be-played music with the highest similarity to the target music piece to the user by analyzing the target music piece.
  • the user can intelligently push a music piece similar to the target music style selected by the user before the user does not select the next to-be-played music piece, so that the user can select the smart music according to the user's preference.
  • control parameters can be played according to different intelligent adjustments of the surrounding environment, so that the robot can play the music in different performance modes in different performance situations, thereby improving the user experience of the listener;
  • the user can provide personalized music recommendation according to the music style data pre-stored by the legitimate user, thereby improving the user experience;
  • the user can intelligently push the music piece similar to the target music piece style selected by the user for the user to select, that is, the user can be intelligently pushed according to the user's preference.
  • the music also improves the user experience.
  • FIG. 3 is a schematic flowchart diagram of still another method for controlling performance of a robot according to an embodiment of the present invention.
  • the performance control method of the robot may include the following steps:
  • the robot constructs a convolutional neural network model, and the convolutional neural network model is used to identify an environment in which the robot is located.
  • the robot constructs a convolutional neural network model.
  • the robot can design a convolutional neural network model consisting of an input layer, three hidden layers, and an output layer, and each layer of nerves.
  • the network corresponds to different weight parameter values.
  • the first layer of the convolutional neural network can find the edge in the image; the second layer can find the graphic formed by the edge found by the first layer.
  • the third layer can look for specific features, such as items with iconic features in the image; each layer passes the image to the next layer, up to the last layer, after finding the target for that layer, and The final output is determined by the combination of all the weight parameter values in the neural network.
  • the robot acquires a large sample image and forms a training sample set according to the massive sample image.
  • the samples in the training sample set can be used to train the neural network model.
  • the robot trains the above convolutional neural network model according to the training sample set, and obtains a trained convolutional neural network model.
  • the convolutional neural network is a feedforward neural network, and the artificial neurons can respond to a surrounding unit in a part of the coverage, and are mainly used for image recognition and processing.
  • the method for controlling the performance of the robot further includes the steps 304 to 306.
  • the steps 304 to 306 refer to the detailed description of the steps 101 to 103 in the first embodiment, and details are not described herein again.
  • the robot controls the robot's built-in camera to capture an image, and uses the trained convolutional neural network model to identify the image to identify the current environment of the robot.
  • the specific manner in which the robot uses the trained convolutional neural network model to identify the image may be: the robot controls the built-in camera of the robot to capture an image.
  • the built-in camera can capture the surrounding environment of the robot.
  • the 360° panoramic image is then preprocessed by the robot, that is, the panoramic image is grayed out, and the gray image is obtained after processing, and then the gray image is input into the trained convolutional neural network model.
  • the grayscale image is sequentially recognized by the five layers of the input layer, the three hidden layers, and the output layer, and the recognition result is outputted by the convolutional neural network model.
  • the current environment of the robot is the auditorium.
  • the robot can capture the panoramic image of the surrounding environment through the built-in camera and identify the panoramic image using the trained convolutional neural network model.
  • the convolutional neural network The model sequentially recognizes that the environment is a concert hall, and then outputs the recognition result that the environment in which the robot is currently located is a concert hall. It can be seen that, in the embodiment of the present invention, the image can be identified by the trained convolutional neural network model to identify the current environment of the robot, and the complex feature extraction process of the image is simplified compared with the traditional image recognition method. The image recognition process also increases the efficiency and accuracy of the recognition environment.
  • the robot adjusts the performance control parameter to a target performance control parameter that matches an environment in which the robot is currently located, and controls the robot to play the target music according to the target performance control parameter.
  • the robot may set a database, and the database prestores a plurality of performance control parameters for playing music, and plays control parameters and environment.
  • the scenes are related, that is, different environmental scenes match different performance control parameters. For example, if the environment scene has a concert hall, a classroom, and a home, then the database can pre-store the performance control parameters that match the performance hall, the performance control parameters that match the classroom, and the performance control parameters that match the home.
  • the performance matching control parameter of the concert hall can be obtained from the database, and the performance control parameter is determined as the target performance control parameter, and the control parameter is played according to the target.
  • the robot is controlled to play the above-mentioned target music. It can be seen that, in the embodiment of the present invention, the performance control parameters of the music piece can be adjusted for different environmental adaptability, so that the robot obtains different interpretation modes for the same piece of music, improves the intelligence of playing the music piece, and makes the performance of the music piece more suitable for the environment.
  • the robot controls the pickup to record the audio signal when the robot plays the target music piece.
  • the robot may control the pickup to record the audio signal when the robot plays the target music in the process of playing the target music, and the starting moment of the audio signal recorded by the pickup is the starting moment of the target music of the robot, and the pickup
  • the end time at which the audio signal is recorded is the end time of the robot playing the target music piece
  • the pickup is a device for collecting the live environment sound and transmitting it to the back end device, which is composed of a microphone (microphone) and an audio amplifying circuit.
  • the robot determines a noise signal from the audio signal, acquires a characteristic parameter of the noise signal, and performs noise reduction processing on the audio signal according to the characteristic parameter to obtain an audio file.
  • the robot after acquiring the audio signal recorded by the pickup, the robot performs signal preprocessing on the audio signal. Specifically, the robot can perform spectrum analysis on the audio signal to determine noise in the audio signal. The signal is obtained, and the characteristic parameter of the noise signal is obtained, and the audio signal is subjected to noise reduction processing according to the characteristic parameter to obtain an audio file.
  • the noise signal may be inversely suppressed according to the phase of the noise signal to obtain an audio file. It can be seen that, in the embodiment of the present invention, the noise signal can be separated from the audio signal to achieve noise reduction on the audio signal, thereby obtaining an audio file with relatively pure sound quality.
  • step 311 The robot detects whether there is a mobile terminal that establishes a communication connection with the robot. If yes, step 312 is performed. Otherwise, if not, the process ends.
  • the robot sends the sharing information to the mobile terminal, where the sharing information includes the audio file.
  • the environment in which the robot is currently located is a high-level concert hall
  • the mobile terminal is a mobile phone of the listener.
  • the audience listened to the target music of the robot in the advanced auditorium, they were very satisfied with the performance version of the target music. I hope that the performance version can be kept in my mobile phone for leisure time, then the audience can enter the advanced performance.
  • the communication connection is established with the robot through the mobile phone; the robot can record and denoise the audio signal when playing the target music piece while playing the target music piece to obtain an audio file, and send the audio file to the listener's mobile phone.
  • the audio file of the target music piece can be subjected to noise reduction processing and then shared to the mobile terminal that establishes a communication connection with the robot, so that the user of the mobile terminal can listen to the audio file at any time, thereby realizing the robot and the mobile terminal.
  • the interaction between users improves the user experience.
  • the robot acquires position coordinates corresponding to the current environment.
  • the robot associates the audio file with the location coordinates to obtain an association relationship between the audio file and the location coordinates.
  • the robot associates the audio file with the position coordinates to obtain an association relationship between the audio file and the position coordinates, that is, establish a correspondence between the audio file and the position coordinates.
  • the robot shares the foregoing relationship to the Internet terminal, so that the Internet terminal recommends the audio file to the mobile terminal of another user when the mobile terminal of the other user exists when the location coordinates are detected.
  • an association relationship between the audio file and the position coordinates may be adopted, so that when the mobile terminal of another user exists at the location coordinate, the audio file is directly recommended to the mobile terminal, thereby realizing Sharing of audio files.
  • the method for controlling the performance of the robot further includes steps 316 to 318.
  • steps 316 to 318 refer to the detailed description of the steps 206 to 208 in the first embodiment, and details are not described herein again.
  • control parameters can be played according to different intelligent adjustments of the surrounding environment, so that the robot can play the music in different performance modes in different performance situations, thereby improving the user experience of the listener;
  • the user can provide personalized music recommendation according to the music style data pre-stored by the legitimate user, thereby improving the user experience;
  • the user can intelligently push the music piece similar to the target music piece style selected by the user for the user to select, that is, the user can be intelligently pushed according to the user's preference. Music, while improving the user experience;
  • the image can be identified by the trained convolutional neural network model to identify the current environment of the robot, which simplifies the image recognition process compared to the traditional image recognition method. It also improves the efficiency and accuracy of the identification environment;
  • the performance control parameters of the music can be adjusted for different environmental adaptability, so that the robot obtains different interpretations of the same piece of music, improves the intelligence of playing the music, and makes the performance of the music more suitable for the environment;
  • the noise signal can be separated from the audio signal to achieve noise reduction on the audio signal, thereby obtaining an audio file with relatively pure sound quality
  • the audio file of the target music piece can be subjected to noise reduction processing and then shared to the mobile terminal that establishes a communication connection with the robot, so that the user of the mobile terminal can listen to the audio file at any time, thereby realizing the interaction between the robot and the user of the mobile terminal.
  • the user's experience is improved; and the relationship between the audio file and the location coordinates can be used to make the above-mentioned audio file be recommended to the mobile terminal immediately after detecting that there are other users' mobile terminals at the location coordinates. To achieve the sharing of audio files.
  • FIG. 4 is a schematic structural diagram of a robot disclosed in an embodiment of the present invention. As shown in FIG. 4, the robot may include:
  • the receiving unit 401 is configured to receive a performance instruction, where the performance instruction includes at least a music name of the target music piece, and provide the music music name to the first acquiring unit 402.
  • the performance instruction received by the receiving unit 401 may be sent by the user through a mobile terminal that is pre-bound with the robot.
  • the form of the performance instruction may be text or voice, which is not limited in the embodiment of the present invention.
  • the user may be a parent, and the mobile terminal pre-bound with the robot may be a parent's mobile phone. Then, when the parent goes out and the child is alone at home, the parent hopes that the robot can play the piano music for the child, thereby cultivating the child's
  • the sentiment, the child's interest in music so parents can send voice messages to the robot through the mobile phone, the voice information includes the music name of the target music that the parents want the robot to play.
  • the robot After receiving the voice information sent by the parent, the robot recognizes the voice information, and plays the piano song corresponding to the music name to the child according to the recognized music name of the target music. It can be seen that, in the embodiment of the present invention, the remote control robot can play the music for the child when the parent goes out, and the real piano performance appreciation is brought to the child, thereby improving the user experience.
  • the first obtaining unit 402 is configured to acquire an original MIDI file that matches the name of the music piece.
  • the first obtaining unit 402 may further generate an original matching the music name by processing the audio file whose music name matches.
  • a MIDI file specifically, the robot can process the audio file matching the music name to obtain time attribute description information of the audio file, and the time attribute description information is a time attribute corresponding to each note in the audio file; Performing a framing process on the audio file according to the preset time length, obtaining a spectrum center of gravity of each audio frame in the audio file, calculating a spectral mean value of the audio file according to the center of gravity of the spectrum, and normalizing the average value of the spectrum to obtain a MIDI file parameter, the MIDI file parameter can be used to describe the pitch information of the audio file; the robot can generate an original MIDI file matching the music name according to the time attribute corresponding to each note in the audio file and the pitch information of the audio file. . It can be seen that, in the embodiment of the present invention, the MIDI file
  • the parsing unit 403 is configured to parse the original MIDI file acquired by the first acquiring unit 402, obtain the performance control parameter of the target music, and trigger the recognition unit 404 to start.
  • the parsing unit 403 parses the original MIDI file, and the MIDI file parameter may be obtained, wherein the MIDI file parameter may include all the notes of the target music and the interval between each note of all the notes and the start of the target music.
  • the note data such as (offset), pitch (frequency), time value (duration), and velocity
  • the MIDI file parameter is the performance control parameter of the above target music.
  • the parsing unit 403 parses the original MIDI file, and can also obtain the overall timbre of the target music.
  • the robot can adjust the overall tone of the target music according to any tone selected by the user, and then adapt the corresponding parameters to the adjusted overall tone.
  • the identification unit 404 is configured to control the built-in camera of the robot to capture an image, and identify an environment in which the robot is currently located according to the image, and trigger the startup adjustment unit 405.
  • the adjusting unit 405 is configured to adjust the performance control parameter obtained by the parsing unit 403 to a target performance control parameter that matches the environment in which the robot is currently located.
  • the performance unit 406 is configured to control the robot performance target music according to the target performance control parameter of the adjustment unit 405.
  • the performance unit 406 controls the robot to play the target music according to the target performance control parameter.
  • the target performance control parameter can be converted into a pulse output signal, that is, the target is read.
  • Playing the note data in the control parameter that is, the interval (offset), pitch (frequency), time value (duration), and velocity of each note in the target song and the start of the target song, and A pulse signal corresponding to each note is formed according to the note data, and pulse signals of different notes are corresponding to different signal output channels, and the signal output channel is associated with the motor of the robot.
  • the performance unit 406 can output a pulse signal in the chronological order of each note, and can drive the motor to start, that is, the robot starts playing the target music.
  • the robot described in FIG. 4 can adjust the performance control parameters according to different intelligent adjustments of the surrounding environment, so that the robot can play the music in different performance modes in different performance situations, thereby improving the user experience of the listener;
  • FIG. 5 is a schematic structural diagram of another robot according to an embodiment of the present invention.
  • the robot shown in FIG. 5 is further optimized by the robot shown in FIG. 4.
  • the robot shown in FIG. 5 may further include:
  • the first detecting unit 407 is configured to detect, after the playing unit 406 controls the robot playing target music according to the target performance control parameter, whether a new performance instruction is received within the preset time period, and the starting time of the preset time period is the target The moment when the music is played.
  • the first detecting unit 407 is pre-configured with a user information base in which a legal face image and/or legal voiceprint data of at least one legal user is pre-stored, and each legal user The favorite music type data, in which the legitimate face image and/or legal voiceprint material of each legal user is bound to the favorite music type data.
  • a legal face image and/or legal voiceprint data of at least one legal user is pre-stored
  • each legal user The favorite music type data in which the legitimate face image and/or legal voiceprint material of each legal user is bound to the favorite music type data.
  • the environment in which the robot is currently located is at home.
  • the first detecting unit 407 can activate the face recognition function and/or the voiceprint recognition function to obtain The user's photo and/or voiceprint, and matching the user's photo and/or voiceprint with the legal face image and/or legal voiceprint data of all legitimate users in the user information library, if the user's photo and/or Or the voiceprint matches the legal face image and/or the legal voiceprint data of any legal user in the user information database, then the first detecting unit 407 can retrieve the favorite user of the matching user in the user information database. The music type data is triggered, and the second acquisition unit 408 is activated.
  • the second acquisition unit 408 analyzes the favorite music type data, and pushes the music to the user that matches the favorite music type data.
  • the robot activates the face recognition function and/or the voiceprint recognition function and the number of users who successfully match the legitimate users in the user information base is greater than one digit, the robot can calculate the favorite music type data of the users. The degree of coincidence to select the push song. It can be seen that, in the embodiment of the present invention, after the legal user is identified through the face and/or the voiceprint, the user can provide the personalized music recommendation according to the music style data pre-stored by the legitimate user, thereby improving the user experience.
  • the second obtaining unit 408 is configured to: when the first detecting unit 407 detects that a new performance instruction is not received within the preset time period, acquire at least one to-be-played music whose similarity with the target music is higher than a preset threshold. And triggering the output unit 409 to start.
  • the output unit 409 is configured to output at least one song name of the music to be played by voice, so that the user selects the target music to be played according to the music name of the at least one music to be played.
  • the robot described in FIG. 5 can adjust the performance control parameters according to the different intelligent adjustments of the surrounding environment, so that the robot can play the music in different performance modes in different performance situations, thereby improving the user experience of the listener;
  • the user can provide personalized music recommendation according to the music style data pre-stored by the legitimate user, thereby improving the user experience;
  • the user can intelligently push the music piece similar to the target music piece style selected by the user for the user to select, that is, the user can be intelligently pushed according to the user's preference.
  • the music also improves the user experience.
  • FIG. 6 is a schematic structural diagram of still another robot disclosed in an embodiment of the present invention. Among them, the robot shown in Fig. 6 is optimized by the robot shown in Fig. 5. Compared with the robot shown in Fig. 5, the robot shown in Fig. 6 further includes:
  • the building unit 410 is configured to construct a convolutional neural network model for identifying an environment in which the robot is located, and triggering the forming unit 411 to start.
  • the building unit 410 constructs a convolutional neural network model.
  • the robot can design a convolutional neural network model consisting of five layers of neural networks: an input layer, three hidden layers, and an output layer, and each The layer neural network corresponds to different weight parameter values.
  • the first layer of the convolutional neural network can find the edge in the image; the second layer can find the graphic formed by the edge found by the first layer.
  • the third layer can look for specific features, such as items with iconic features in the image; each layer passes the image to the next layer, up to the last layer, after finding the target for that layer, and The final output is determined by the combination of all weight parameter values in the neural network.
  • the forming unit 411 is configured to acquire a massive sample image and form a training sample set according to the massive sample image.
  • the training unit 412 is configured to train the convolutional neural network model constructed by the constructing unit 410 according to the training sample set formed by the forming unit 411 to obtain a trained convolutional neural network model.
  • the identification unit 404 is specifically configured to identify the image by using the convolutional neural network model trained by the training unit 412 to identify the environment in which the robot is currently located.
  • the specific unit of the recognition unit 404 that uses the trained convolutional neural network model trained by the training unit 412 may be: controlling the built-in camera of the robot to capture an image.
  • the built-in camera can capture a 360° panoramic image of the surrounding environment of the robot, and then pre-processing the panoramic image, that is, grayening the panoramic image, obtaining a grayscale image after processing, and then inputting the grayscale image into the trained convolutional nerve
  • the network model allows the grayscale image to be sequentially identified by the five layers of the input layer, the three hidden layers, and the output layer, and the recognition result is outputted by the convolutional neural network model.
  • the recognition unit 404 can capture a panoramic image of the surrounding environment through the built-in camera, and identify the panoramic image by the recognition unit 404 using the trained convolutional neural network model.
  • the convolutional neural network model sequentially recognizes that the environment is a concert hall by layering, and then the recognition unit 404 outputs the recognition result that the environment in which the robot is currently located is a concert hall.
  • the image can be identified by the trained convolutional neural network model to identify the current environment of the robot, and the complex feature extraction process of the image is simplified compared with the traditional image recognition method.
  • the image recognition process also increases the efficiency and accuracy of the recognition environment.
  • the robot shown in FIG. 6 may further include:
  • the recording unit 413 is configured to control the audio signal when the player records the target music piece in the process of playing the target music piece by the performance unit 406, and provide the audio signal to the processing unit 414.
  • the processing unit 414 is configured to determine a noise signal from the audio signal, acquire a characteristic parameter of the noise signal, perform noise reduction processing on the audio signal according to the feature parameter, obtain an audio file, and trigger the second detecting unit 415 to start.
  • the processing unit 414 after acquiring the audio signal recorded by the recording unit 413, performs signal preprocessing on the audio signal. Specifically, the processing unit 414 may perform spectrum analysis on the audio signal to determine the a noise signal in the audio signal, and acquiring a characteristic parameter of the noise signal, and performing noise reduction processing on the audio signal according to the characteristic parameter to obtain an audio file. Optionally, after determining the noise signal in the audio signal, the processing unit 414 may also perform inverse suppression processing on the noise signal according to the phase of the noise signal to obtain an audio file. It can be seen that, in the embodiment of the present invention, the noise signal can be separated from the audio signal to achieve noise reduction on the audio signal, thereby obtaining an audio file with relatively pure sound quality.
  • the second detecting unit 415 is configured to detect whether there is a mobile terminal that establishes a communication connection with the robot.
  • the sending unit 416 is configured to: when the second detecting unit 415 detects that there is a mobile terminal that establishes a communication connection with the robot, send the sharing information to the mobile terminal, where the sharing information includes the audio file obtained by the processing unit 414.
  • the first obtaining unit 402 is further configured to acquire location coordinates corresponding to the current environment.
  • the robot shown in FIG. 6 further includes:
  • the association unit 417 is configured to associate the audio file obtained by the processing unit 414 with the location coordinates acquired by the first acquiring unit 402, obtain an association relationship between the audio file and the location coordinates, and provide the association relationship to the sharing. Unit 418.
  • the sharing unit 418 is configured to share the association relationship to the Internet terminal, so that the Internet terminal recommends the audio file to the mobile terminal of another user when the mobile terminal of the other user exists when the location coordinates are detected.
  • the robot described in FIG. 6 can adjust the performance control parameters according to different intelligent adjustments of the surrounding environment, so that the robot can play the music in different performance modes in different performance situations, thereby improving the user experience of the listener;
  • the user can provide personalized music recommendation according to the music style data pre-stored by the legitimate user, thereby improving the user experience;
  • the user can intelligently push the music piece similar to the target music piece style selected by the user for the user to select, that is, the user can be intelligently pushed according to the user's preference. Music, while improving the user experience;
  • the image can be identified by the trained convolutional neural network model to identify the current environment of the robot, which simplifies the image recognition process compared to the traditional image recognition method. It also improves the efficiency and accuracy of the identification environment;
  • the performance control parameters of the music can be adjusted for different environmental adaptability, so that the robot obtains different interpretations of the same piece of music, improves the intelligence of playing the music, and makes the performance of the music more suitable for the environment;
  • the noise signal can be separated from the audio signal to achieve noise reduction on the audio signal, thereby obtaining an audio file with relatively pure sound quality
  • the audio file of the target music piece can be subjected to noise reduction processing and then shared to the mobile terminal that establishes a communication connection with the robot, so that the user of the mobile terminal can listen to the audio file at any time, thereby realizing the interaction between the robot and the user of the mobile terminal.
  • the user's experience is improved; and the relationship between the audio file and the location coordinates can be used to make the above-mentioned audio file be recommended to the mobile terminal immediately after detecting that there are other users' mobile terminals at the location coordinates. To achieve the sharing of audio files.
  • FIG. 7 is a schematic structural diagram of another robot disclosed in an embodiment of the present invention. As shown in Figure 7, the robot can include:
  • a memory 701 storing executable program code
  • processor 702 coupled to the memory 701;
  • the processor 702 calls the executable program code stored in the memory 701 to execute the performance control method of the robot of any of FIGS. 1 to 3.
  • the embodiment of the invention discloses a computer readable storage medium storing a computer program, wherein the computer program causes the computer to execute the performance control method of the robot of any of FIGS. 1 to 3.
  • B corresponding to A means that B is associated with A, and B can be determined according to A.
  • determining B according to A does not mean that B is only determined based on A, and that B can also be determined based on A and/or other information.
  • the units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
  • each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
  • the above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.
  • the above integrated units if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer-accessible memory.
  • the technical solution of the present invention which is essential or contributes to the prior art, or all or part of the technical solution, may be embodied in the form of a software product stored in a memory.
  • a number of requests are included to cause a computer device (which may be a personal computer, server or network device, etc., and in particular a processor in a computer device) to perform some or all of the steps of the above-described methods of various embodiments of the present invention.
  • ROM Read-Only Memory
  • RAM Random Access Memory
  • PROM Programmable Read-Only Memory
  • EPROM Erasable Programmable Read Only Memory
  • OTPROM One-Time Programmable Read-Only Memory
  • EEPROM Electronically-Erasable Programmable Read-Only Memory
  • CD-ROM Compact Disc Read-Only Memory

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Acoustics & Sound (AREA)
  • Manipulator (AREA)
  • Toys (AREA)

Abstract

本发明涉及人工智能技术领域，具体涉及一种机器人的演奏控制方法及机器人，包括：机器人接收演奏指令，该演奏指令至少包括目标乐曲的乐曲名称，获取并解析与该乐曲名称相匹配的原始MIDI文件，得到目标乐曲的演奏控制参数，以及控制机器人的内置摄像头拍摄图像，并根据图像识别机器人当前所处的环境，然后将演奏控制参数调整为与机器人当前所处的环境相匹配的目标演奏控制参数，根据目标演奏控制参数控制机器人演奏目标乐曲。可见，本发明实施例，能够使机器人在不同的演奏场合下采用不同的演奏方式，从而提高聆听者的用户体验。

Description

一种机器人的演奏控制方法及机器人 技术领域
本发明涉及人工智能技术领域，具体涉及一种机器人的演奏控制方法及机器人。
背景技术
日新月异的科技发展给音乐领域注入了许多新鲜元素，逐渐改变了人们编写、表演、制作甚至欣赏音乐的方式，其中，得益于人工智能技术的飞速发展，机器人也可以像演奏家一样，为人们演奏乐曲，给人们带来不一样的听觉体验。对于机器人来说，可以通过机器学习的方法学习不同的演奏家演奏同一乐曲的版本，由此获得演奏该乐曲的演奏方式，演奏方式包括演奏该乐曲的演奏速度、演奏力度等。然而，相比起演奏家可以在不同的演奏场合采用不同的演奏方式演奏同一首乐曲，机器人在演奏同一首乐曲时只能通过固定的演奏方式进行演奏，机器人的演奏方式过于单一。
发明内容
本发明实施例公开一种机器人的演奏控制方法及机器人，能够使机器人在不同的演奏场合下采用不同的演奏方式，从而提高聆听者的用户体验。
所述机器人接收演奏指令，所述演奏指令至少包括目标乐曲的乐曲名称；
所述机器人获取与所述乐曲名称相匹配的原始MIDI文件；
所述机器人解析所述原始MIDI文件，得到所述目标乐曲的演奏控制参数；
所述机器人控制所述机器人的内置摄像头拍摄图像，并根据所述图像识别所述机器人当前所处的环境；
所述机器人将所述演奏控制参数调整为与所述机器人当前所处的环境相匹配的目标演奏控制参数，并根据所述目标演奏控制参数控制所述机器人演奏所述目标乐曲。
作为一种可选的实施方式，在本发明实施例第一方面中，所述机器人根据所述目标演奏控制参数控制所述机器人演奏所述目标乐曲之后，所述方法还包括：
所述机器人检测预设时间段内是否接收到新的演奏指令，所述预设时间段的起始时刻是所述目标乐曲演奏结束的时刻；
如果否，则所述机器人获取与所述目标乐曲相似度高于预设阈值的至少一首待演奏乐曲；
所述机器人通过语音输出至少一首所述待演奏乐曲的乐曲名称，以使用户根据至少一首所述待演奏乐曲的乐曲名称选择目标待演奏乐曲。
作为一种可选的实施方式，在本发明实施例第一方面中，所述方法还包括：
所述机器人构建卷积神经网络模型，所述卷积神经网络模型用于识别所述机器人所处的环境；
所述机器人获取海量样本图像，并根据所述海量样本图像形成训练样本集；
所述机器人根据所述训练样本集对所述卷积神经网络模型进行训练，得到训练好的卷积神经网络模型；
所述机器人根据所述图像识别所述机器人当前所处的环境，包括：
所述机器人使用所述训练好的卷积神经网络模型对所述图像进行识别，以识别出所述机器人当前所处的环境。
作为一种可选的实施方式，在本发明实施例第一方面中，所述方法还包括：
在演奏所述目标乐曲的过程中，所述机器人控制拾音器记录所述机器人演奏所述目标乐曲时的音频信号；
所述机器人从所述音频信号中确定出噪声信号，获取所述噪声信号的特征参数，并根据所述特征参数对所述音频信号进行降噪处理，得到音频文件；
所述机器人检测是否存在与所述机器人建立通信连接的移动终端；
如果存在，则所述机器人向所述移动终端发送分享信息，所述分享信息包括所述音频文件。
作为一种可选的实施方式，在本发明实施例第一方面中，所述方法还包括：
所述机器人获取当前所处的环境对应的位置坐标；
所述机器人将所述音频文件与所述位置坐标进行关联，得到所述音频文件与所述位置坐标的关联关系；
所述机器人将所述关联关系分享至互联网终端，以使所述互联网终端在检测到所述位置坐标处存在其它用户的移动终端时，向所述其它用户的移动终端推荐所述音频文件。
本发明实施例第二方面公开了一种机器人，所述机器人包括：
接收单元，用于接收演奏指令，所述演奏指令至少包括目标乐曲的乐曲名称；
第一获取单元，用于获取与所述乐曲名称相匹配的原始MIDI文件；
解析单元，用于解析所述原始MIDI文件，得到所述目标乐曲的演奏控制参数；
识别单元，用于控制所述机器人的内置摄像头拍摄图像，并根据所述图像识别所述机器人当前所处的环境；
调整单元，用于将所述演奏控制参数调整为与所述机器人当前所处的环境相匹配的目标演奏控制参数；
演奏单元，用于根据所述目标演奏控制参数控制所述机器人演奏所述目标乐曲。
作为一种可选的实施方式，在本发明实施例第二方面中，所述机器人还包括：
第一检测单元，用于在所述演奏单元根据所述目标演奏控制参数控制所述机器人演奏所述目标乐曲之后，检测预设时间段内是否接收到新的演奏指令，所述预设时间段的起始时刻是所述目标乐曲演奏结束的时刻；
第二获取单元，用于在所述第一检测单元检测出所述预设时间段内没有接收到所述新的演奏指令时，获取与所述目标乐曲相似度高于预设阈值的至少一首待演奏乐曲；
输出单元，用于通过语音输出至少一首所述待演奏乐曲的乐曲名称，以使用户根据至少一首所述待演奏乐曲的乐曲名称选择目标待演奏乐曲。
作为一种可选的实施方式，在本发明实施例第二方面中，所述机器人还包括：
构建单元，用于构建卷积神经网络模型，所述卷积神经网络模型用于识别所述机器人所处的环境；
形成单元，用于获取海量样本图像，并根据所述海量样本图像形成训练样本集；
训练单元，用于根据所述训练样本集对所述卷积神经网络模型进行训练，得到训练好的卷积神经网络模型；
所述识别单元，具体用于使用所述训练单元得到的所述训练好的卷积神经网络模型对所述图像进行识别，以识别出所述机器人当前所处的环境。
作为一种可选的实施方式，在本发明实施例第二方面中，所述机器人还包括：
记录单元，用于在所述演奏单元演奏所述目标乐曲的过程中，控制拾音器记录所述机器人演奏所述目标乐曲时的音频信号；
处理单元，用于从所述音频信号中确定出噪声信号，获取所述噪声信号的特征参数，并根据所述特征参数对所述音频信号进行降噪处理，得到音频文件；
第二检测单元，用于检测是否存在与所述机器人建立通信连接的移动终端；
发送单元，用于在所述第二检测单元检测出存在与所述机器人建立通信连接的移动终端时，向所述移动终端发送分享信息，所述分享信息包括所述音频文件。
作为一种可选的实施方式，在本发明实施例第二方面中，所述第一获取单元，还用于获取当前所处的环境对应的位置坐标；
所述机器人还包括：
关联单元，用于将所述音频文件与所述位置坐标进行关联，得到所述音频文件与所述位置坐标的关联关系；
分享单元，用于将所述关联关系分享至互联网终端，以使所述互联网终端在检测到所述位置坐标处存在其它用户的移动终端时，向所述其它用户的移动终端推荐所述音频文件。
本发明实施例第三方面公开另一种机器人，包括：
存储有可执行程序代码的存储器；
与所述存储器耦合的处理器；
所述处理器调用所述存储器中存储的所述可执行程序代码，执行本发明实施例第一方面公开的一种机器人的演奏控制方法。
本发明实施例第四方面公开一种计算机可读存储介质，其存储计算机程序，其中，所述计算机程序使得计算机执行本发明实施例第一方面公开的一种机器人的演奏控制方法。
本发明实施例第五方面公开一种计算机程序产品，当所述计算机程序产品在计算机上运行时，使得所述计算机执行第一方面的任意一种方法的部分或全部步骤。
本发明实施例第六方面公开一种应用发布平台，所述应用发布平台用于发布计算机程序产品，其中，当所述计算机程序产品在计算机上运行时，使得所述计算机执行第一方面的任意一种方法的部分或全部步骤。
与现有技术相比，本发明实施例具有以下有益效果：
本发明实施例中，机器人接收演奏指令，该演奏指令至少包括目标乐曲的乐曲名称，获取并解析与该乐曲名称相匹配的原始MIDI文件，得到目标乐曲的演奏控制参数，以及控制机器人的内置摄像头拍摄图像，并根据图像识别机器人当前所处的环境，然后将演奏控制参数调整为与机器人当前所处的环境相匹配的目 标演奏控制参数，根据目标演奏控制参数控制机器人演奏目标乐曲。可见，实施本发明实施例，能够根据周围环境的不同调整演奏控制参数，以使机器人在不同的演奏场合下采用不同的演奏方式演奏乐曲，从而提高了聆听者的用户体验。
附图说明
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本发明实施例公开的一种机器人的演奏控制方法的流程示意图；
图2是本发明实施例公开的另一种机器人的演奏控制方法的流程示意图；
图3是本发明实施例公开的又一种机器人的演奏控制方法的流程示意图；
图4是本发明实施例公开的一种机器人的结构示意图；
图5是本发明实施例公开的另一种机器人的结构示意图；
图6是本发明实施例公开的又一种机器人的结构示意图；
图7是本发明实施例公开的又一种机器人的结构示意图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
需要说明的是，本发明实施例及附图中的术语“包括”和“具有”以及它们任何变形，意图在于覆盖不排他的包含。例如包含了一系列步骤或单元的过程、方法、系统、产品或设备没有限定于已列出的步骤或单元，而是可选地还包括没有列出的步骤或单元，或可选地还包括对于这些过程、方法、产品或设备固有的其它步骤或单元。
本发明实施例公开一种机器人的演奏控制方法及机器人，能够使机器人在不同的演奏场合下采用不同的演奏方式，从而提高聆听者的用户体验。以下分别进行详细说明。
实施例一
请参阅图1，图1是本发明实施例公开的一种机器人的演奏控制方法的流程示意图。如图1所示，该机器人的演奏控制方法可以包括以下操作：
101、机器人接收演奏指令，该演奏指令至少包括目标乐曲的乐曲名称。
本发明实施例中，该演奏指令可以是用户通过与该机器人预先绑定的移动终端发送的，该演奏指令的形式可以是文本也可以是语音，本发明实施例不做限定。举例来说，用户可以是家长，与该机器人预先绑定的移动终端可以是家长的手机，那么，当家长外出而孩子独自在家时，家长希望机器人能够为孩子演奏钢琴曲，以此陶冶孩子的情操、培养孩子对音乐的兴趣，所以家长可以通过手机向机器人发送语音信息，该语音信息包括家长希望机器人演奏的目标乐曲的乐曲名称。机器人接收到家长发送的语音信息之后，对该语音信息进行识别，并根据识别出的 目标乐曲的乐曲名称为孩子演奏该乐曲名称对应的钢琴曲。可见，本发明实施例，可以实现家长外出时远程控制机器人为孩子演奏乐曲，为孩子带来真实的钢琴演奏欣赏，提高了用户体验。
102、机器人获取与上述乐曲名称相匹配的原始MIDI文件。
本发明实施例中，MIDI(Musical Instrument Digital Interface)，也称作乐器数字接口，是一个工业标准的电子通信协议，为电子乐器等演奏设备(如合成器)定义各种音符或弹奏码，容许电子乐器、电脑、手机或其它的舞台演出配备彼此连接，调整和同步，得以即时交换演奏数据。对于MIDI文件来说，MIDI文件主要分为为文件头信息(Header Chunk)以及音轨信息(Track Chunk)两个部分，文件头信息主要储存MIDI文件中的基本信息，如：每个四分音符被切割为多少点数(Tick)、音轨信息区块数、文件格式代号等等，而音轨信息则用来储存一连串的MIDI信息，依照不同的文件格式代号，MIDI信息可以用来表示单一音轨(文件格式代号0)或者是多个音轨(文件格式代号1)的MIDI信息，本发明实施例不做限定。
作为一种可选的实施方式，在获取与上述乐曲名称相匹配的原始MIDI文件之前，机器人可以通过处理上述乐曲名称相匹配的音频文件，生成与上述乐曲名称相匹配的原始MIDI文件，具体来说，机器人可以对与上述乐曲名称相匹配的音频文件进行处理，得到该音频文件的时间属性描述信息，该时间属性描述信息是音频文件中每一个音符对应的时间属性；然后按照预设时间长度对该音频文件进行分帧处理，得到该音频文件中各个音频帧的频谱重心，并根据该频谱重心计算音频文件的频谱均值，以及对该频谱均值进行归一化处理，得到MIDI文件参数，该MIDI参数可用于描述该音频文件的音准信息；机器人可以根据该音频文件中每一个音符对应的时间属性以及该音频文件的音准信息生成与上述乐曲名称相匹配的原始MIDI文件。可见，本发明实施例，可以通过音频文件计算并生成与该音频文件对应的MIDI文件，提高了音频处理的效率，进而也提升了音频处理的智能性。
103、机器人解析原始MIDI文件，得到上述目标乐曲的演奏控制参数。
本发明实施例中，机器人解析原始MIDI文件，可以得到MIDI文件参数，其中，该MIDI文件参数可以包括上述目标乐曲的所有音符以及所有音符中每个音符与目标乐曲起始处的间隔时间(偏移量)、音高(频率)、时值(持续时长)以及力度等音符数据，该MIDI文件参数即为上述目标乐曲的演奏控制参数。
本发明实施例中，可选的，机器人解析原始MIDI文件，还可以得到上述目标乐曲的整体音色。机器人可以依照用户选择的任意音色来调整该目标乐曲的整体音色，然后对调整后的整体音色进行对应参数的适应化处理。
104、机器人控制该机器人的内置摄像头拍摄图像，并根据该图像识别机器人当前所处的环境。
本发明实施例中，机器人可以控制内置摄像头拍摄图像，优选的，机器人可以控制内置摄像头拍摄全景图像，即机器人可以预设拍摄范围以及设置多个全景对焦位置，并通过内置摄像头拍摄与上述多个全景对焦位置对应的多个对焦图像，并由这多个对焦图像生成全景图像与预设拍摄范围相对应的全景图像。
105、机器人将上述演奏控制参数调整为与该机器人当前所处的环境相匹配 的目标演奏控制参数，并根据目标演奏控制参数控制该机器人演奏上述目标乐曲。
本发明实施例中，根据目标演奏控制参数控制该机器人演奏上述目标乐曲，具体来说，机器人演奏目标乐曲时，可以将目标演奏控制参数转换成脉冲输出信号，即读入目标演奏控制参数中的音符数据，该音符数据即目标乐曲中每个音符与目标乐曲起始处的间隔时间(偏移量)、音高(频率)、时值(持续时长)以及力度等，并根据该音符数据形成每个音符对应的脉冲信号，并将不同的音符的脉冲信号对应不同的信号输出通道，而信号输出通道与机器人的电机建立关联关系。举例来说，在机器人通过钢琴演奏乐曲时，机器人可以按照每个音符的时间顺序输出脉冲信号，可以驱动电机启动，即机器人开始演奏目标乐曲。
可见，通过图1所描述的方法，能够根据周围环境的不同智能的调整演奏控制参数，以使机器人在不同的演奏场合下采用不同的演奏方式演奏乐曲，从而提高了聆听者的用户体验；
此外，可以实现家长外出时远程控制机器人为孩子演奏乐曲，为孩子带来真实的钢琴演奏欣赏，提高了用户体验；以及，可以通过音频文件计算并生成与该音频文件对应的MIDI文件，提高了音频处理的效率，进而也提升了音频处理的智能性。
实施例二
请参阅图2，图2是本发明实施例公开的另一种机器人的演奏控制方法的流程示意图。如图2所示，该机器人的演奏控制方法可以包括以下步骤：
本发明实施例中，该机器人的演奏控制方法还包括步骤201～205，针对步骤201～205的描述，请参照实施例一中针对步骤101～105的详细描述，本发明实施例不再赘述。
206、机器人检测预设时间段内是否接收到新的演奏指令，如果是，则根据新的演奏指令演奏一首新的乐曲，反之，如果否，则执行步骤207。
本发明实施例中，上述预设时间段的起始时刻是目标乐曲演奏结束的时刻。
作为一种可选的实施方式，机器人预设有用户信息库，该用户信息库中预存有至少一个合法用户的合法人脸图像和/或合法声纹资料，以及每个合法用户的喜爱乐曲类型资料，其中，每个合法用户的合法人脸图像和/或合法声纹资料与喜爱乐曲类型资料相绑定。举例来说，机器人当前所处的环境是家里，当机器人检测预设时间段内未接收到新的演奏指令时，机器人可以启动人脸识别功能和/或声纹识别功能，获取用户的照片和/或声纹，并将该用户的照片和/或声纹与用户信息库中所有合法用户的合法人脸图像和/或合法声纹资料进行匹配，如果该用户的照片和/或声纹与用户信息库中任意一位合法用户的合法人脸图像和/或合法声纹资料相匹配，那么，机器人可以调取用户信息库中该匹配成功的合法用户的喜爱乐曲类型资料，并分析该喜爱乐曲类型资料，向用户推送与该喜爱乐曲类型资料相匹配的乐曲。特别的，当机器人启动人脸识别功能和/或声纹识别功能识别出的、并且与用户信息库中合法用户匹配成功的用户数量大于一位时，机器人可以通过计算这些用户的喜爱乐曲类型资料的重合度来选择推送乐曲。可见，本发明实施例，可以在通过人脸和/或声纹识别出合法用户之后，根据合法用户预存的乐曲风格资料，为用户提供个性化的乐曲推荐，提高了用户的体验度。
207、机器人获取与上述目标乐曲相似度高于预设阈值的至少一首待演奏乐曲。
208、机器人通过语音输出至少一首待演奏乐曲的乐曲名称，以使用户根据至少一首待演奏乐曲的乐曲名称选择目标待演奏乐曲。
针对步骤206～步骤208，机器人可以预先设置预设时间段，该预设时间段的起始时刻是上一首乐曲演奏结束的结束时刻。那么，在一首乐曲演奏完毕之后，如果机器人在预设时间段内接收到新的演奏指令，机器人可以根据新的演奏指令中包括的新的乐曲名称演奏该新的乐曲；如果机器人在预设时间段内未接收到新的演奏指令，机器人可以通过分析上述目标乐曲，向用户推送与上述目标乐曲相似度极高的待演奏乐曲供用户选择。可见，本发明实施例，可以在用户没有选择下一首待演奏乐曲情况下，智能的向用户推送与用户之前选择的目标乐曲音乐风格相似的乐曲供用户选择，即可以根据用户喜好智能的推送该用户感兴趣的乐曲，同时提高了用户体验度。
可见，通过图2所描述的方法，能够根据周围环境的不同智能的调整演奏控制参数，以使机器人在不同的演奏场合下采用不同的演奏方式演奏乐曲，从而提高了聆听者的用户体验；
此外，可以实现家长外出时远程控制机器人为孩子演奏乐曲，为孩子带来真实的钢琴演奏欣赏，提高了用户体验；以及，可以通过音频文件计算并生成与该音频文件对应的MIDI文件，提高了音频处理的效率，进而也提升了音频处理的智能性；
此外，可以在通过人脸和/或声纹识别出合法用户之后，根据合法用户预存的乐曲风格资料，为用户提供个性化的乐曲推荐，提高了用户的体验度；
此外，可以在用户没有选择下一首待演奏乐曲情况下，智能的向用户推送与用户之前选择的目标乐曲音乐风格相似的乐曲供用户选择，即可以根据用户喜好智能的推送该用户感兴趣的乐曲，同时提高了用户体验度。
实施例三
请参阅图3，图3是本发明实施例公开的又一种机器人的演奏控制方法的流程示意图。如图3所示，该机器人的演奏控制方法可以包括以下步骤：
301、机器人构建卷积神经网络模型，该卷积神经网络模型用于识别机器人所处的环境。
本发明实施例中，机器人构建卷积神经网络模型，具体来说，机器人可以设计由输入层、三个隐层和输出层这五层神经网络构成的卷积神经网络模型，并且每一层神经网络对应不同的权重参数值。举例来说，机器人在通过该卷积神经网络模型识别图像的内容时，卷积神经网络的第一层可以寻找该图像中的边缘；第二层可以寻找第一层找到的边缘所组成的图形，例如长方形或者圆形；第三层可以寻找特定特征，例如图像中具有标志性特征的物品；每一层在寻找到该层的目标之后将图像传递至下一层，直至最后一层，而最终的输出结果由神经网络中所有的权重参数值共同决定。
302、机器人获取海量样本图像，并根据海量样本图像形成训练样本集。
本发明实施例中，训练样本集中的样本可以用于训练神经网络模型。
303、机器人根据训练样本集对上述卷积神经网络模型进行训练，得到训练 好的卷积神经网络模型。
本发明实施例中，卷积神经网络是一种前馈神经网络，它的人工神经元可以响应一部分覆盖范围内的周围单元，主要用于图像的识别、处理方面。
本发明实施例中，该机器人的演奏控制方法还包括步骤304～306，针对步骤304～306的描述，请参照实施例一中针对步骤101～103的详细描述，本发明实施例不再赘述。
307、机器人控制该机器人的内置摄像头拍摄图像，并使用上述训练好的卷积神经网络模型对该图像进行识别，以识别出机器人当前所处的环境。
本发明实施例中，机器人使用上述训练好的卷积神经网络模型对该图像进行识别的具体方式可以为：机器人控制该机器人的内置摄像头拍摄图像，优选的，该内置摄像头可以拍摄机器人周围环境的360°的全景图像，然后机器人可以对该全景图像进行预处理，即对该全景图像进行灰度化处理，处理后得到灰度图像，然后将该灰度图像输入训练好的卷积神经网络模型，使该灰度图像依次经过输入层、三个隐层和输出层这五层神经网络的识别，由该卷积神经网络模型识别之后输出识别结果。举例来说，机器人当前所处的环境是演奏厅，那么，机器人可以通过内置摄像头拍摄周围环境的全景图像，并使用训练好的卷积神经网络模型对该全景图像进行识别，该卷积神经网络模型通过分层依次识别出该环境是演奏厅，那么输出识别结果即为该机器人当前所处的环境是演奏厅。可见，本发明实施例，可以通过训练好的卷积神经网络模型对图像进行识别，以识别出机器人当前所处的环境，相比起传统图像识别方式对图像的复杂的特征提取过程，简化了图像识别过程，从而也提高了识别环境的效率和准确性。
308、机器人将上述演奏控制参数调整为与该机器人当前所处的环境相匹配的目标演奏控制参数，并根据目标演奏控制参数控制该机器人演奏上述目标乐曲。
本发明实施例中，由于不同的环境的空间位置环境、噪音强度、听众人数等因素不同，机器人可以设置数据库，该数据库预存有多个用于演奏乐曲的演奏控制参数，并且演奏控制参数与环境场景进行关联，即不同的环境场景匹配不同的演奏控制参数。举例来说，环境场景有演奏厅、教室以及家等，那么，数据库中可以预存有与演奏厅相匹配的演奏控制参数、与教室相匹配的演奏控制参数以及与家相匹配的演奏控制参数，当机器人根据步骤308识别出机器人当前所处的环境是演奏厅时，可以从数据库中获取该演奏厅相匹配演奏控制参数，并将该演奏控制参数确定为目标演奏控制参数，并根据目标演奏控制参数控制该机器人演奏上述目标乐曲。可见，本发明实施例，可以针对不同的环境适应性调整乐曲的演奏控制参数，使得机器人获得对同一首乐曲的不同演绎方式，提高了演奏乐曲的智能性，使乐曲的演奏更贴合环境。
309、在演奏上述目标乐曲的过程中，机器人控制拾音器记录该机器人演奏上述目标乐曲时的音频信号。
本发明实施例中，机器人可以在演奏目标乐曲的过程中，控制拾音器记录该机器人演奏上述目标乐曲时的音频信号，拾音器记录该音频信号的起始时刻是机器人演奏目标乐曲的起始时刻，拾音器记录该音频信号的终止时刻是机器人演奏目标乐曲的终止时刻，其中，拾音器是用来采集现场环境声音再传送到后端设备 的一个器件，它是由咪头(麦克风)和音频放大电路构成。
310、机器人从上述音频信号中确定出噪声信号，获取噪声信号的特征参数，并根据特征参数对该音频信号进行降噪处理，得到音频文件。
本发明实施例中，在获取到拾音器记录的音频信号之后，机器人对该音频信号进行信号预处理，具体来说，就是机器人可以对该音频信号进行频谱分析，以确定出该音频信号中的噪声信号，并获取噪声信号的特征参数，根据该特征参数对该音频信号进行降噪处理，得到音频文件。可选的，机器人确定出音频信号中的噪声信号之后，也可以根据噪声信号相位，对噪声信号进行反相抑制处理，得到音频文件。可见，本发明实施例，可以从音频信号中分离出噪声信号，实现对音频信号的降噪，从而获得音质较为纯净的音频文件。
311、机器人检测是否存在与该机器人建立通信连接的移动终端，如果存在，则执行步骤312，反之，如果不存在，结束本流程。
312、机器人向移动终端发送分享信息，该分享信息包括上述音频文件。
针对步骤309～步骤312，举例来说，机器人当前所处的环境是高级演奏厅，移动终端是听众的手机。当听众在高级演奏厅聆听机器人演奏的目标乐曲之后，对此次目标乐曲的演奏版本十分满意，希望能够将该演奏版本保留在自己的手机里以便闲暇时间欣赏，那么，听众可以在进入高级演奏厅之前通过手机与机器人建立通信连接；机器人可以在演奏目标乐曲的同时对演奏该目标乐曲时的音频信号进行记录以及降噪处理得到音频文件，并将该音频文件发送到该听众的手机上。可见，本发明实施例，可以将演奏目标乐曲的音频文件进行降噪处理之后分享给与机器人建立通信连接的移动终端，使移动终端的用户可以随时聆听该音频文件，实现了机器人与移动终端的用户之间的交互，提高了用户的体验度。
313、机器人获取当前所处的环境对应的位置坐标。
314、机器人将上述音频文件与上述位置坐标进行关联，得到音频文件与位置坐标的关联关系。
本发明实施例中，机器人将上述音频文件与上述位置坐标进行关联，得到音频文件与位置坐标的关联关系，即建立音频文件与位置坐标之间的对应关系。
315、机器人将上述关联关系分享至互联网终端，以使互联网终端在检测到位置坐标处存在其它用户的移动终端时，向其它用户的移动终端推荐上述音频文件。
针对步骤313～步骤315，可以通过音频文件与位置坐标之间形成的关联关系，使得一旦检测到该位置坐标处存在其他用户的移动终端时，立即向该移动终端推荐上述的音频文件，从而实现音频文件的共享。
其中，该机器人的演奏控制方法还包括步骤316～318，针对步骤316～318的描述，请参照实施例一中针对步骤206～208的详细描述，本发明实施例不再赘述。
可见，通过图3所描述的方法，能够根据周围环境的不同智能的调整演奏控制参数，以使机器人在不同的演奏场合下采用不同的演奏方式演奏乐曲，从而提高了聆听者的用户体验；
此外，可以实现家长外出时远程控制机器人为孩子演奏乐曲，为孩子带来真实的钢琴演奏欣赏，提高了用户体验；以及，可以通过音频文件计算并生成与该音频文件对应的MIDI文件，提高了音频处理的效率，进而也提升了音频处理的 智能性；
此外，可以在通过人脸和/或声纹识别出合法用户之后，根据合法用户预存的乐曲风格资料，为用户提供个性化的乐曲推荐，提高了用户的体验度；
此外，可以在用户没有选择下一首待演奏乐曲情况下，智能的向用户推送与用户之前选择的目标乐曲音乐风格相似的乐曲供用户选择，即可以根据用户喜好智能的推送该用户感兴趣的乐曲，同时提高了用户体验度；
此外，可以通过训练好的卷积神经网络模型对图像进行识别，以识别出机器人当前所处的环境，相比起传统图像识别方式对图像的复杂的特征提取过程，简化了图像识别过程，从而也提高了识别环境的效率和准确性；
此外，可以针对不同的环境适应性调整乐曲的演奏控制参数，使得机器人获得对同一首乐曲的不同演绎方式，提高了演奏乐曲的智能性，使乐曲的演奏更贴合环境；
此外，可以从音频信号中分离出噪声信号，实现对音频信号的降噪，从而获得音质较为纯净的音频文件；
此外，可以将演奏目标乐曲的音频文件进行降噪处理之后分享给与机器人建立通信连接的移动终端，使移动终端的用户可以随时聆听该音频文件，实现了机器人与移动终端的用户之间的交互，提高了用户的体验度；以及，可以通过音频文件与位置坐标之间形成的关联关系，使得一旦检测到该位置坐标处存在其他用户的移动终端时，立即向该移动终端推荐上述的音频文件，从而实现音频文件的共享。
实施例四
请参阅图4，图4是本发明实施例公开的一种机器人的结构示意图。如图4所示，该机器人可以包括：
接收单元401，用于接收演奏指令，该演奏指令至少包括目标乐曲的乐曲名称，并将该乐曲名称提供给第一获取单元402。
本发明实施例中，接收单元401接收的演奏指令可以是用户通过与该机器人预先绑定的移动终端发送的，该演奏指令的形式可以是文本也可以是语音，本发明实施例不做限定。举例来说，用户可以是家长，与该机器人预先绑定的移动终端可以是家长的手机，那么，当家长外出而孩子独自在家时，家长希望机器人能够为孩子演奏钢琴曲，以此陶冶孩子的情操、培养孩子对音乐的兴趣，所以家长可以通过手机向机器人发送语音信息，该语音信息包括家长希望机器人演奏的目标乐曲的乐曲名称。机器人接收到家长发送的语音信息之后，对该语音信息进行识别，并根据识别出的目标乐曲的乐曲名称为孩子演奏该乐曲名称对应的钢琴曲。可见，本发明实施例，可以实现家长外出时远程控制机器人为孩子演奏乐曲，为孩子带来真实的钢琴演奏欣赏，提高了用户体验。
第一获取单元402，用于获取与上述乐曲名称相匹配的原始MIDI文件。
作为一种可选的实施方式，第一获取单元402在获取与上述乐曲名称相匹配的原始MIDI文件之前，还可以通过处理上述乐曲名称相匹配的音频文件，生成与上述乐曲名称相匹配的原始MIDI文件，具体来说，机器人可以对与上述乐曲名称相匹配的音频文件进行处理，得到该音频文件的时间属性描述信息，该时间属性描述信息是音频文件中每一个音符对应的时间属性；然后按照预设时间长度 对该音频文件进行分帧处理，得到该音频文件中各个音频帧的频谱重心，并根据该频谱重心计算音频文件的频谱均值，以及对该频谱均值进行归一化处理，得到MIDI文件参数，该MIDI文件参数可用于描述该音频文件的音准信息；机器人可以根据该音频文件中每一个音符对应的时间属性以及该音频文件的音准信息生成与上述乐曲名称相匹配的原始MIDI文件。可见，本发明实施例，可以通过音频文件计算并生成与该音频文件对应的MIDI文件，提高了音频处理的效率，进而也提升了音频处理的智能性。
解析单元403，用于解析上述第一获取单元402获取到的原始MIDI文件，得到上述目标乐曲的演奏控制参数，并触发识别单元404启动。
本发明实施例中，解析单元403解析原始MIDI文件，可以得到MIDI文件参数，其中，该MIDI文件参数可以包括上述目标乐曲的所有音符以及所有音符中每个音符与目标乐曲起始处的间隔时间(偏移量)、音高(频率)、时值(持续时长)以及力度等音符数据，该MIDI文件参数即为上述目标乐曲的演奏控制参数。
本发明实施例中，可选的，解析单元403解析原始MIDI文件，还可以得到上述目标乐曲的整体音色。机器人可以依照用户选择的任意音色来调整该目标乐曲的整体音色，然后对调整后的整体音色进行对应参数的适应化处理。
识别单元404，用于控制机器人的内置摄像头拍摄图像，并根据该图像识别机器人当前所处的环境，并触发启动调整单元405。
调整单元405，用于将上述解析单元403得到的演奏控制参数调整为与该机器人当前所处的环境相匹配的目标演奏控制参数。
演奏单元406，用于根据上述调整单元405的目标演奏控制参数控制该机器人演奏目标乐曲。
本发明实施例中，演奏单元406根据目标演奏控制参数控制该机器人演奏上述目标乐曲，具体来说，演奏单元406演奏目标乐曲时，可以将目标演奏控制参数转换成脉冲输出信号，即读入目标演奏控制参数中的音符数据，该音符数据即目标乐曲中每个音符与目标乐曲起始处的间隔时间(偏移量)、音高(频率)、时值(持续时长)以及力度等，并根据该音符数据形成每个音符对应的脉冲信号，并将不同的音符的脉冲信号对应不同的信号输出通道，而信号输出通道与机器人的电机建立关联关系。举例来说，在机器人通过钢琴演奏乐曲时，演奏单元406可以按照每个音符的时间顺序输出脉冲信号，可以驱动电机启动，即机器人开始演奏目标乐曲。
可见，通过图4所描述的机器人，能够根据周围环境的不同智能的调整演奏控制参数，以使机器人在不同的演奏场合下采用不同的演奏方式演奏乐曲，从而提高了聆听者的用户体验；
此外，可以实现家长外出时远程控制机器人为孩子演奏乐曲，为孩子带来真实的钢琴演奏欣赏，提高了用户体验；以及，可以通过音频文件计算并生成与该音频文件对应的MIDI文件，提高了音频处理的效率，进而也提升了音频处理的智能性。
实施例五
请参阅图5，图5是本发明实施例提供的另一种机器人的结构示意图，其中， 图5所示的机器人是由图4所示的机器人进一步进行优化得到的。与图4所示的机器人相比较，图5所示的机器人还可以包括：
第一检测单元407，用于在上述演奏单元406根据目标演奏控制参数控制机器人演奏目标乐曲之后，检测预设时间段内是否接收到新的演奏指令，该预设时间段的起始时刻是目标乐曲演奏结束的时刻。
作为一种可选的实施方式，第一检测单元407预设有用户信息库，该用户信息库中预存有至少一个合法用户的合法人脸图像和/或合法声纹资料，以及每个合法用户的喜爱乐曲类型资料，其中，每个合法用户的合法人脸图像和/或合法声纹资料与喜爱乐曲类型资料相绑定。举例来说，机器人当前所处的环境是家里，当机器人检测预设时间段内未接收到新的演奏指令时，第一检测单元407可以启动人脸识别功能和/或声纹识别功能，获取用户的照片和/或声纹，并将该用户的照片和/或声纹与用户信息库中所有合法用户的合法人脸图像和/或合法声纹资料进行匹配，如果该用户的照片和/或声纹与用户信息库中任意一位合法用户的合法人脸图像和/或合法声纹资料相匹配，那么，第一检测单元407可以调取用户信息库中该匹配成功的合法用户的喜爱乐曲类型资料，并触发第二获取单元408启动，由第二获取单元408分析该喜爱乐曲类型资料，向用户推送与该喜爱乐曲类型资料相匹配的乐曲。特别的，当机器人启动人脸识别功能和/或声纹识别功能识别出的、并且与用户信息库中合法用户匹配成功的用户数量大于一位时，机器人可以通过计算这些用户的喜爱乐曲类型资料的重合度来选择推送乐曲。可见，本发明实施例，可以在通过人脸和/或声纹识别出合法用户之后，根据合法用户预存的乐曲风格资料，为用户提供个性化的乐曲推荐，提高了用户的体验度。
第二获取单元408，用于在上述第一检测单元407检测出预设时间段内没有接收到新的演奏指令时，获取与目标乐曲相似度高于预设阈值的至少一首待演奏乐曲，并触发输出单元409启动。
输出单元409，用于通过语音输出至少一首待演奏乐曲的乐曲名称，以使用户根据至少一首待演奏乐曲的乐曲名称选择目标待演奏乐曲。
可见，通过图5所描述的机器人，能够根据周围环境的不同智能的调整演奏控制参数，以使机器人在不同的演奏场合下采用不同的演奏方式演奏乐曲，从而提高了聆听者的用户体验；
此外，可以实现家长外出时远程控制机器人为孩子演奏乐曲，为孩子带来真实的钢琴演奏欣赏，提高了用户体验；以及，可以通过音频文件计算并生成与该音频文件对应的MIDI文件，提高了音频处理的效率，进而也提升了音频处理的智能性；
此外，可以在通过人脸和/或声纹识别出合法用户之后，根据合法用户预存的乐曲风格资料，为用户提供个性化的乐曲推荐，提高了用户的体验度；
此外，可以在用户没有选择下一首待演奏乐曲情况下，智能的向用户推送与用户之前选择的目标乐曲音乐风格相似的乐曲供用户选择，即可以根据用户喜好智能的推送该用户感兴趣的乐曲，同时提高了用户体验度。
实施例六
请参阅图6，图6是本发明实施例公开的又一种机器人的结构示意图。其中，图6所示的机器人是由图5所示的机器人进行优化得到的。与图5所示的机器人 相比较，图6所示的机器人还包括：
构建单元410，用于构建卷积神经网络模型，该卷积神经网络模型用于识别机器人所处的环境，并触发形成单元411启动。
本发明实施例中，构建单元410构建卷积神经网络模型，具体来说，机器人可以设计由输入层、三个隐层和输出层这五层神经网络构成的卷积神经网络模型，并且每一层神经网络对应不同的权重参数值。举例来说，机器人在通过该卷积神经网络模型识别图像的内容时，卷积神经网络的第一层可以寻找该图像中的边缘；第二层可以寻找第一层找到的边缘所组成的图形，例如长方形或者圆形；第三层可以寻找特定特征，例如图像中具有标志性特征的物品；每一层在寻找到该层的目标之后将图像传递至下一层，直至最后一层，而最终的输出由神经网络中所有的权重参数值共同决定。
形成单元411，用于获取海量样本图像，并根据海量样本图像形成训练样本集。
训练单元412，用于根据上述形成单元411形成的训练样本集对上述构建单元410构建的卷积神经网络模型进行训练，得到训练好的卷积神经网络模型。
上述识别单元404，具体用于使用上述训练单元412训练好的卷积神经网络模型对图像进行识别，以识别出该机器人当前所处的环境。
本发明实施例中，识别单元404使用上述训练单元412训练好的卷积神经网络模型对该图像进行识别的具体方式可以为：控制该机器人的内置摄像头拍摄图像，优选的，该内置摄像头可以拍摄机器人周围环境的360°的全景图像，然后对该全景图像进行预处理，即对该全景图像进行灰度化处理，处理后得到灰度图像，然后将该灰度图像输入训练好的卷积神经网络模型，使该灰度图像依次经过输入层、三个隐层和输出层这五层神经网络的识别，由该卷积神经网络模型识别之后输出识别结果。举例来说，机器人当前所处的环境是演奏厅，那么，识别单元404可以通过内置摄像头拍摄周围环境的全景图像，并通过识别单元404使用训练好的卷积神经网络模型对该全景图像进行识别，该卷积神经网络模型通过分层依次识别出该环境是演奏厅，那么识别单元404输出识别结果即为该机器人当前所处的环境是演奏厅。可见，本发明实施例，可以通过训练好的卷积神经网络模型对图像进行识别，以识别出机器人当前所处的环境，相比起传统图像识别方式对图像的复杂的特征提取过程，简化了图像识别过程，从而也提高了识别环境的效率和准确性。
作为一种可选的实施方式，图6所示的机器人还可以包括：
记录单元413，用于在上述演奏单元406演奏目标乐曲的过程中，控制拾音器记录机器人演奏目标乐曲时的音频信号，并将该音频信号提供给处理单元414。
处理单元414，用于从上述音频信号中确定出噪声信号，获取该噪声信号的特征参数，并根据特征参数对上述音频信号进行降噪处理，得到音频文件，并触发第二检测单元415启动。
本发明实施例中，在获取记录单元413记录的音频信号之后，处理单元414对该音频信号进行信号预处理，具体来说，就是处理单元414可以对该音频信号进行频谱分析，以确定出该音频信号中的噪声信号，并获取噪声信号的特征参数，根据该特征参数对该音频信号进行降噪处理，得到音频文件。可选的，处理单元 414在确定出音频信号中的噪声信号之后，也可以根据噪声信号相位，对噪声信号进行反相抑制处理，得到音频文件。可见，本发明实施例，可以从音频信号中分离出噪声信号，实现对音频信号的降噪，从而获得音质较为纯净的音频文件。
第二检测单元415，用于检测是否存在与该机器人建立通信连接的移动终端。
发送单元416，用于在上述第二检测单元415检测出存在与机器人建立通信连接的移动终端时，向移动终端发送分享信息，该分享信息包括上述处理单元414得到的音频文件。
上述第一获取单元402，还用于获取当前所处的环境对应的位置坐标。
可选的，图6所示的机器人还包括：
关联单元417，用于将上述处理单元414得到的音频文件与上述第一获取单元402获取到的位置坐标进行关联，得到该音频文件与该位置坐标的关联关系，并将该关联关系提供给分享单元418。
分享单元418，用于将关联关系分享至互联网终端，以使互联网终端在检测到该位置坐标处存在其它用户的移动终端时，向其它用户的移动终端推荐上述音频文件。
可见，通过图6所描述的机器人，能够根据周围环境的不同智能的调整演奏控制参数，以使机器人在不同的演奏场合下采用不同的演奏方式演奏乐曲，从而提高了聆听者的用户体验；
此外，可以实现家长外出时远程控制机器人为孩子演奏乐曲，为孩子带来真实的钢琴演奏欣赏，提高了用户体验；以及，可以通过音频文件计算并生成与该音频文件对应的MIDI文件，提高了音频处理的效率，进而也提升了音频处理的智能性；
此外，可以在通过人脸和/或声纹识别出合法用户之后，根据合法用户预存的乐曲风格资料，为用户提供个性化的乐曲推荐，提高了用户的体验度；
此外，可以在用户没有选择下一首待演奏乐曲情况下，智能的向用户推送与用户之前选择的目标乐曲音乐风格相似的乐曲供用户选择，即可以根据用户喜好智能的推送该用户感兴趣的乐曲，同时提高了用户体验度；
此外，可以通过训练好的卷积神经网络模型对图像进行识别，以识别出机器人当前所处的环境，相比起传统图像识别方式对图像的复杂的特征提取过程，简化了图像识别过程，从而也提高了识别环境的效率和准确性；
此外，可以针对不同的环境适应性调整乐曲的演奏控制参数，使得机器人获得对同一首乐曲的不同演绎方式，提高了演奏乐曲的智能性，使乐曲的演奏更贴合环境；
此外，可以从音频信号中分离出噪声信号，实现对音频信号的降噪，从而获得音质较为纯净的音频文件；
此外，可以将演奏目标乐曲的音频文件进行降噪处理之后分享给与机器人建立通信连接的移动终端，使移动终端的用户可以随时聆听该音频文件，实现了机器人与移动终端的用户之间的交互，提高了用户的体验度；以及，可以通过音频文件与位置坐标之间形成的关联关系，使得一旦检测到该位置坐标处存在其他用户的移动终端时，立即向该移动终端推荐上述的音频文件，从而实现音频文件的共享。
请参阅图7，图7是本发明实施例公开的另一种机器人的结构示意图。如图7所示，该机器人可以包括：
存储有可执行程序代码的存储器701；
与存储器701耦合的处理器702；
其中，处理器702调用存储器701中存储的可执行程序代码，执行图1～图3任意一种机器人的演奏控制方法。
本发明实施例公开一种计算机可读存储介质，其存储计算机程序，其中，该计算机程序使得计算机执行图1～图3任意一种机器人的演奏控制方法。
应理解，说明书通篇中提到的“一个实施例”或“一实施例”意味着与实施例有关的特定特征、结构或特性包括在本发明的至少一个实施例中。因此，在整个说明书各处出现的“在一个实施例中”或“在一实施例中”未必一定指相同的实施例。此外，这些特定特征、结构或特性可以以任意适合的方式结合在一个或多个实施例中。本领域技术人员也应该知悉，说明书中所描述的实施例均属于可选实施例，所涉及的动作和模块并不一定是本发明所必须的。
在本发明的各种实施例中，应理解，上述各过程的序号的大小并不意味着执行顺序的必然先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本发明实施例的实施过程构成任何限定。
在本发明的各种实施例中，应理解，本文中术语“和/或”，仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B这三种情况。另外，本文中字符“/”，一般表示前后关联对象是一种“或”的关系。
在本发明所提供的实施例中，应理解，“与A相应的B”表示B与A相关联，根据A可以确定B。但还应理解，根据A确定B并不意味着仅仅根据A确定B，还可以根据A和/或其他信息确定B。
上述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可位于一个地方，或者也可以分布到多个网络单元上。可根据实际的需要选择其中的部分或全部单元来实现本实施例方案的目的。
另外，在本发明各实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。
上述集成的单元若以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可获取的存储器中。基于这样的理解，本发明的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的全部或者部分，可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储器中，包括若干请求用以使得一台计算机设备(可以为个人计算机、服务器或者网络设备等，具体可以是计算机设备中的处理器)执行本发明的各个实施例上述方法的部分或全部步骤。
本领域普通技术人员可以理解上述实施例的各种方法中的全部或部分步骤是可以通过程序来指令相关的硬件来完成，该程序可以存储于一计算机可读存储 介质中，存储介质包括只读存储器(Read-Only Memory，ROM)、随机存储器(Random Access Memory，RAM)、可编程只读存储器(Programmable Read-only Memory，PROM)、可擦除可编程只读存储器(Erasable Programmable Read Only Memory，EPROM)、一次可编程只读存储器(One-time Programmable Read-Only Memory，OTPROM)、电子抹除式可复写只读存储器(Electrically-Erasable Programmable Read-Only Memory，EEPROM)、只读光盘(Compact Disc Read-Only Memory，CD-ROM)或其他光盘存储器、磁盘存储器、磁带存储器、或者能够用于携带或存储数据的计算机可读的任何其他介质。
以上对本发明实施例公开的一种机器人的演奏控制方法及机器人进行了详细介绍，本文中应用了具体个例对本发明的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本发明的方法及其核心思想；同时，对于本领域的一般技术人员，依据本发明的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本发明的限制。

Claims (10)

  1. 一种机器人的演奏控制方法，其特征在于，所述方法包括：
    所述机器人接收演奏指令，所述演奏指令至少包括目标乐曲的乐曲名称；
    所述机器人获取与所述乐曲名称相匹配的原始MIDI文件；
    所述机器人解析所述原始MIDI文件，得到所述目标乐曲的演奏控制参数；
    所述机器人控制所述机器人的内置摄像头拍摄图像，并根据所述图像识别所述机器人当前所处的环境；
    所述机器人将所述演奏控制参数调整为与所述机器人当前所处的环境相匹配的目标演奏控制参数，并根据所述目标演奏控制参数控制所述机器人演奏所述目标乐曲。
  2. 根据权利要求1所述的方法，其特征在于，所述机器人根据所述目标演奏控制参数控制所述机器人演奏所述目标乐曲之后，所述方法还包括：
    所述机器人检测预设时间段内是否接收到新的演奏指令，所述预设时间段的起始时刻是所述目标乐曲演奏结束的时刻；
    如果否，则所述机器人获取与所述目标乐曲相似度高于预设阈值的至少一首待演奏乐曲；
    所述机器人通过语音输出至少一首所述待演奏乐曲的乐曲名称，以使用户根据至少一首所述待演奏乐曲的乐曲名称选择目标待演奏乐曲。
  3. 根据权利要求1或2所述的方法，其特征在于，所述方法还包括：
    所述机器人构建卷积神经网络模型，所述卷积神经网络模型用于识别所述机器人所处的环境；
    所述机器人获取海量样本图像，并根据所述海量样本图像形成训练样本集；
    所述机器人根据所述训练样本集对所述卷积神经网络模型进行训练，得到训练好的卷积神经网络模型；
    所述机器人根据所述图像识别所述机器人当前所处的环境，包括：
    所述机器人使用所述训练好的卷积神经网络模型对所述图像进行识别，以识别出所述机器人当前所处的环境。
  4. 根据权利要求1-3任一项所述的方法，其特征在于，所述方法还包括：
    在演奏所述目标乐曲的过程中，所述机器人控制拾音器记录所述机器人演奏所述目标乐曲时的音频信号；
    所述机器人从所述音频信号中确定出噪声信号，获取所述噪声信号的特征参数，并根据所述特征参数对所述音频信号进行降噪处理，得到音频文件；
    所述机器人检测是否存在与所述机器人建立通信连接的移动终端；
    如果存在，则所述机器人向所述移动终端发送分享信息，所述分享信息包括所述音频文件。
  5. 根据权利要求4所述的方法，其特征在于，所述方法还包括：
    所述机器人获取当前所处的环境对应的位置坐标；
    所述机器人将所述音频文件与所述位置坐标进行关联，得到所述音频文件与所述位置坐标的关联关系；
    所述机器人将所述关联关系分享至互联网终端，以使所述互联网终端在检测到所述位置坐标处存在其它用户的移动终端时，向所述其它用户的移动终端推荐 所述音频文件。
  6. 一种机器人，其特征在于，包括：
    接收单元，用于接收演奏指令，所述演奏指令至少包括目标乐曲的乐曲名称；
    第一获取单元，用于获取与所述乐曲名称相匹配的原始MIDI文件；
    解析单元，用于解析所述原始MIDI文件，得到所述目标乐曲的演奏控制参数；
    识别单元，用于控制所述机器人的内置摄像头拍摄图像，并根据所述图像识别所述机器人当前所处的环境；
    调整单元，用于将所述演奏控制参数调整为与所述机器人当前所处的环境相匹配的目标演奏控制参数；
    演奏单元，用于根据所述目标演奏控制参数控制所述机器人演奏所述目标乐曲。
  7. 根据权利要求6所述的机器人，其特征在于，所述机器人还包括：
    第一检测单元，用于在所述演奏单元根据所述目标演奏控制参数控制所述机器人演奏所述目标乐曲之后，检测预设时间段内是否接收到新的演奏指令，所述预设时间段的起始时刻是所述目标乐曲演奏结束的时刻；
    第二获取单元，用于在所述第一检测单元检测出所述预设时间段内没有接收到所述新的演奏指令时，获取与所述目标乐曲相似度高于预设阈值的至少一首待演奏乐曲；
    输出单元，用于通过语音输出至少一首所述待演奏乐曲的乐曲名称，以使用户根据至少一首所述待演奏乐曲的乐曲名称选择目标待演奏乐曲。
  8. 根据权利要求6或7所述的机器人，其特征在于，所述机器人还包括：
    构建单元，用于构建卷积神经网络模型，所述卷积神经网络模型用于识别所述机器人所处的环境；
    形成单元，用于获取海量样本图像，并根据所述海量样本图像形成训练样本集；
    训练单元，用于根据所述训练样本集对所述卷积神经网络模型进行训练，得到训练好的卷积神经网络模型；
    所述识别单元，具体用于使用所述训练单元得到的所述训练好的卷积神经网络模型对所述图像进行识别，以识别出所述机器人当前所处的环境。
  9. 根据权利要求6-8任一项所述的机器人，其特征在于，所述机器人还包括：
    记录单元，用于在所述演奏单元演奏所述目标乐曲的过程中，控制拾音器记录所述机器人演奏所述目标乐曲时的音频信号；
    处理单元，用于从所述音频信号中确定出噪声信号，获取所述噪声信号的特征参数，并根据所述特征参数对所述音频信号进行降噪处理，得到音频文件；
    第二检测单元，用于检测是否存在与所述机器人建立通信连接的移动终端；
    发送单元，用于在所述第二检测单元检测出存在与所述机器人建立通信连接的移动终端时，向所述移动终端发送分享信息，所述分享信息包括所述音频文件。
  10. 根据权利要求9所述的机器人，其特征在于，所述第一获取单元，还用于获取当前所处的环境对应的位置坐标；
    所述机器人还包括：
    关联单元，用于将所述音频文件与所述位置坐标进行关联，得到所述音频文件与所述位置坐标的关联关系；
    分享单元，用于将所述关联关系分享至互联网终端，以使所述互联网终端在检测到所述位置坐标处存在其它用户的移动终端时，向所述其它用户的移动终端推荐所述音频文件。
PCT/CN2017/117681 2017-12-12 2017-12-21 一种机器人的演奏控制方法及机器人 WO2019114015A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711322453.6A CN108053815A (zh) 2017-12-12 2017-12-12 一种机器人的演奏控制方法及机器人
CN201711322453.6 2017-12-12

Publications (1)

Publication Number Publication Date
WO2019114015A1 true WO2019114015A1 (zh) 2019-06-20

Family

ID=62131964

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/117681 WO2019114015A1 (zh) 2017-12-12 2017-12-21 一种机器人的演奏控制方法及机器人

Country Status (2)

Country Link
CN (1) CN108053815A (zh)
WO (1) WO2019114015A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113008234A (zh) * 2021-02-09 2021-06-22 北京智能佳科技有限公司 基于室内定位装置的群体协作系统
CN114495959A (zh) * 2021-12-14 2022-05-13 科大讯飞股份有限公司 语音增强方法、装置和系统

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108922315A (zh) * 2018-08-21 2018-11-30 福州大学厦门工艺美术学院 一种吉他辅助练习装置
CN110136677B (zh) * 2019-03-28 2022-03-15 深圳市芒果未来科技有限公司 乐音控制方法及相关产品
CN110827789B (zh) * 2019-10-12 2023-05-23 平安科技（深圳）有限公司 音乐生成方法、电子装置及计算机可读存储介质
CN116394277B (zh) * 2023-06-08 2023-08-25 之江实验室 一种仿人钢琴演奏机器人

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102314917A (zh) * 2010-07-01 2012-01-11 北京中星微电子有限公司 一种影音文件的播放方法及播放装置
CN103051794A (zh) * 2012-12-18 2013-04-17 广东欧珀移动通信有限公司 动态设置移动终端音效的方法和装置
CN103714836A (zh) * 2012-09-29 2014-04-09 联想(北京)有限公司 一种播放音频信息的方法及电子设备
CN104135705A (zh) * 2014-06-24 2014-11-05 惠州Tcl移动通信有限公司 一种根据不同场景模式自动调整多媒体音量的方法及系统
CN104883642A (zh) * 2015-03-27 2015-09-02 成都上生活网络科技有限公司 一种音效调节方法
CN105761742A (zh) * 2014-12-19 2016-07-13 闫辉 一种音乐播放笔帽
US20160371371A1 (en) * 2013-03-15 2016-12-22 Futurewei Technologies, Inc. Motion-based music recommendation for mobile devices
US20170060927A1 (en) * 2015-09-01 2017-03-02 International Business Machines Corporation Song selection using a heart rate change and a facial expression monitored with a camera
CN107332994A (zh) * 2017-06-29 2017-11-07 深圳传音控股有限公司 一种调音效果自匹配方法及系统

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3945429B2 (ja) * 2003-03-19 2007-07-18 ヤマハ株式会社 自動演奏装置および自動演奏プログラム
CN101271318A (zh) * 2007-03-21 2008-09-24 明基电通股份有限公司 交互式家庭娱乐机器人及相关控制方法
CN103200300B (zh) * 2012-01-10 2016-02-10 宇龙计算机通信科技(深圳)有限公司 终端及资源分享方法
CN103095814B (zh) * 2012-12-31 2016-01-20 合一网络技术(北京)有限公司 一种基于信息网络的声音驱动的近距离分享方法、客户端、服务器及系统
CN104978973B (zh) * 2014-10-22 2019-08-13 广州酷狗计算机科技有限公司 一种音频处理方法及装置
CN106774814A (zh) * 2015-11-20 2017-05-31 北京奇虎科技有限公司 智能穿戴设备运动中的歌曲分享方法和装置
CN106446930B (zh) * 2016-06-28 2019-11-22 沈阳工业大学 基于深层卷积神经网络的机器人工作场景识别方法

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102314917A (zh) * 2010-07-01 2012-01-11 北京中星微电子有限公司 一种影音文件的播放方法及播放装置
CN103714836A (zh) * 2012-09-29 2014-04-09 联想(北京)有限公司 一种播放音频信息的方法及电子设备
CN103051794A (zh) * 2012-12-18 2013-04-17 广东欧珀移动通信有限公司 动态设置移动终端音效的方法和装置
US20160371371A1 (en) * 2013-03-15 2016-12-22 Futurewei Technologies, Inc. Motion-based music recommendation for mobile devices
CN104135705A (zh) * 2014-06-24 2014-11-05 惠州Tcl移动通信有限公司 一种根据不同场景模式自动调整多媒体音量的方法及系统
CN105761742A (zh) * 2014-12-19 2016-07-13 闫辉 一种音乐播放笔帽
CN104883642A (zh) * 2015-03-27 2015-09-02 成都上生活网络科技有限公司 一种音效调节方法
US20170060927A1 (en) * 2015-09-01 2017-03-02 International Business Machines Corporation Song selection using a heart rate change and a facial expression monitored with a camera
CN107332994A (zh) * 2017-06-29 2017-11-07 深圳传音控股有限公司 一种调音效果自匹配方法及系统

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113008234A (zh) * 2021-02-09 2021-06-22 北京智能佳科技有限公司 基于室内定位装置的群体协作系统
CN113008234B (zh) * 2021-02-09 2023-04-18 北京智能佳科技有限公司 基于室内定位装置的群体协作系统
CN114495959A (zh) * 2021-12-14 2022-05-13 科大讯飞股份有限公司 语音增强方法、装置和系统

Also Published As

Publication number Publication date
CN108053815A (zh) 2018-05-18

Similar Documents

Publication Publication Date Title
WO2019114015A1 (zh) 一种机器人的演奏控制方法及机器人
JP7404067B2 (ja) ライブ音楽実演のマルチメディア・コンテンツのネットワーク・ベースの処理および配送
US11551669B2 (en) Locally distributed keyword detection
JP7317850B2 (ja) オーディオ合成方法、並びにそのコンピュータプログラム、コンピュータ装置及びコンピュータ装置により構成されるコンピュータシステム
US20210035572A1 (en) Locally distributed keyword detection
RU2513772C2 (ru) Система и способ автоматического создания атмосферы, подходящей для общественной обстановки и настроя в окружающей среде
CN108829765A (zh) 一种信息查询方法、装置、计算机设备及存储介质
JP2020034895A (ja) 応答方法及び装置
CN102959544A (zh) 用于同步媒体的方法和系统
JP7453712B2 (ja) オーディオ再生方法、装置、コンピュータ可読記憶媒体及び電子機器
CN106488311B (zh) 音效调整方法及用户终端
CN110675886A (zh) 音频信号处理方法、装置、电子设备及存储介质
JP6977323B2 (ja) 歌唱音声の出力方法、音声応答システム、及びプログラム
US11771866B2 (en) Locally distributed keyword detection
CN113691909B (zh) 具有音频处理推荐的数字音频工作站
JP2022092032A (ja) 歌唱合成システム及び歌唱合成方法
US11574627B2 (en) Masking systems and methods
CN112422999B (zh) 直播内容处理方法及计算机设备
WO2022041192A1 (zh) 语音消息处理方法、设备及即时通信客户端
TWI377559B (en) Singing system with situation sound effect and method thereof
CN111627417B (zh) 播放语音的方法、装置及电子设备
WO2022041177A1 (zh) 通信消息处理方法、设备及即时通信客户端
WO2023010949A1 (zh) 一种音频数据的处理方法及装置
WO2021210338A1 (ja) 再生制御方法、制御システムおよびプログラム
US20230252979A1 (en) Gatekeeping for voice intent processing

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17934492

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17934492

Country of ref document: EP

Kind code of ref document: A1