According to the data of official website, the academic system of "Music Artificial Intelligence and Music Information Technology" is three years, and the applicants must be candidates majoring in computer, intelligence and electronic information.
As for the recommendation list, except 1' s book Fundamentals of Music Theory, the other four recommendation lists are all related to artificial intelligence theory, namely Data Structure and Algorithm, Introduction to Signals and Systems, Artificial Intelligence: A Modern Method and Neural Network and Machine Learning.
Because "Music Artificial Intelligence and Music Information Technology" is an interdisciplinary major, the interview will not only assess the professional ability of the subject, but also assess the musical ability of the candidates-playing a certain instrument or simply singing.
At present, the three joint training instructors that have been finalized for this major are:
Yu feng
Dean of Central Conservatory of Music, professor, doctoral supervisor, talent leaders of "Ten Thousand Talents Program" and "Four Groups". President of the China Command Society, deputy director of the National Teaching Committee for Postgraduates of Art Majors, member of the 10th National Committee of the China Federation of Literary and Art Circles, and enjoying special government allowance from the State Council.
Sun maosong
Professor Tsinghua University, doctoral supervisor, executive vice president of Tsinghua University Institute of Artificial Intelligence, former director of the computer department and secretary of the Party Committee, vice chairman of the Steering Committee for Teaching Informatization and Teaching Method Innovation of the Ministry of Education, and member of the Ninth National Committee of China Association for Science and Technology. His main research fields are natural language processing, artificial intelligence, machine learning and computational pedagogy. Chief scientist of the National 973 Program and chief expert of major projects of the National Social Science Fund. 20 17, leading the development of "nine songs" artificial intelligence ancient poetry writing system.
Wu xihong
Professor Peking University is a doctoral supervisor and an outstanding talent of the Ministry of Education in the new century. Eecs, vice president, director of intelligent science department and director of speech and hearing research center, is devoted to the research in the fields of machine auditory computing theory, speech information processing, natural language understanding, music intelligence, etc. He has presided over more than 40 national, provincial and ministerial projects, won more than 10 invention patents authorized by the state, and published more than 200 academic papers. He has made great achievements in the field of intelligent music creation and arrangement.
Candidates who are interested in enrolling in this major must complete online registration (website:/) from March 1 day to March 15, 1965, and the examination will be held in the Central Conservatory of Music in May this year.
For more details, please click:
/xwyhd/xsjd/20 19s/20 1903/t 20 19030 1 _ 53856 . html
know each other
Is there an early warning for professional opening?
If you have been paying attention to the development of the Central Conservatory of Music, you will not be surprised by the opening of this major.
As early as May last year, the Central Conservatory of Music signed a cooperation agreement with the School of Information Computing and Engineering of Indiana University, which is famous for its innovative interdisciplinary research, to jointly build an Information Philharmonic Orchestra laboratory-the so-called "Information Philharmonic Orchestra", which refers to a musical artificial intelligence accompaniment system, invented by Christopher Raphael, director of the Music Informatics Laboratory of the School of Information Computing and Engineering of Indiana University.
The biggest feature of the system is that it can comprehensively interpret and calculate the music itself and the musician's feelings by mathematical methods. Through continuous active learning, the template of orchestra accompaniment and concerto is formed, which is closer to the individual performance needs of musicians and provides musicians with more flexible performance opportunities.
After signing the contract, after more than half a year of intensive preparations, the two sides jointly held the first special concert in China, AI Night Concert, accompanied by artificial intelligence on October 26th last year. 12 outstanding soloists from Central Conservatory of Music and Information Philharmonic Orchestra performed 12 Chinese and foreign works of various genres together.
It is worth mentioning that China music "Capriccio of the Great Wall" is accompanied by artificial intelligence, which is the first collision between artificial intelligence technology and Chinese folk music.
The picture is from official website, Central Conservatory of Music.
Professor Ryan, Dean of the Central Conservatory of Music, said in his concert speech: "This is a far-reaching concert, and the whole music industry in China will enter an era of' artificial intelligence' from it, which greatly improves the information level of the whole music industry, especially the music education industry. The combination of artificial intelligence technology and music art specialty will realize the leap-forward development of the whole industry and will surely become a model of music industry industrialization. 」
Full performance video of "Ai Night Concert":
/index.php？ option=weixin，dianbodetail & ampid=35 14
The enthusiasm for domestic scientific research is growing.
In addition to the Central Conservatory of Music, Xinghai Conservatory of Music and Central University for Nationalities have also made achievements in trying artificial intelligence+music.
From May 65438 to June last year, the joint laboratory of music artificial intelligence aided orchestral teaching, which was jointly established by the Orchestral Department of Xinghai Conservatory of Music and the Music Informatics Laboratory of the School of Information Computing and Engineering of Indiana University, was officially launched. The two sides will cooperate to introduce the music artificial intelligence assisted orchestral teaching system into daily teaching.
It is understood that the system allows students to listen to the complete music accompaniment of professional orchestras at any time in their daily professional practice, and at the same time, convert the synthetic performance audio of themselves and the orchestra into highly structured, visual, retrievable and comparable music data to be brought to class for discussion with professional teachers. For professional teachers, the system can compare students' professional learning vertically and horizontally, obtain students' first-hand information, and improve teaching contents and methods.
The picture comes from the official account of WeChat of Xinghai Conservatory of Music.
On February 7th last year, 65438+, the signing ceremony of "Artificial Intelligence Music Joint Laboratory" jointly organized by Minzu University of China and Ping An Technology was held in Zhixingtang, Minzu University of China. This cooperation aims to give full play to their respective advantages and realize the idea of artificial intelligence music creation from appreciation stage to professional stage to expert stage through joint research and development.
Song Min, member of the Standing Committee of the Party Committee of the Central University for Nationalities and vice president, said at the unveiling ceremony that artificial intelligence has been included in the national plan and entered the stage of gradual implementation, and it is constantly being combined with various fields, which will undoubtedly lead the development of all walks of life in the future. She hopes that the two sides will give full play to their respective advantages through the platform of the laboratory, improve the discipline construction level and music creation level of the Central University for Nationalities, promote the construction of the "four centers" in Beijing, especially the cultural center, and actively help China's excellent music culture to go abroad.
The picture is from official website, Central University for Nationalities.
In addition, the 6th China Conference on Sound and Music Technology (CSMT), co-founded by Fudan University and Tsinghua University, has been exporting academic views on the cross-cutting field of sound and music technology for China since 20 13, enriching the research results in the field of artificial intelligence and music in China.
Taking the 20 18 meeting as an example, the topics of the paper include:
musical acoustics
Instrumental acoustics/vocal acoustics/psychoacoustics and electro-acoustics/space music acoustics, etc.
Signal processing of sound and music
Sound signal processing/music signal processing in industry, agriculture, animal husbandry, aquaculture, geography, environment and other fields.
Computer listening
Content analysis, understanding and modeling of sound and music/audio and music information retrieval/classification, annotation, emotional calculation, recommendation, etc. /Application of artificial intelligence in music computing/Application of music computing in entertainment, education, ocean, medicine, equipment, military affairs, information security and other fields.
Audio information security
Powerful audio watermarking/audio authentication/audio forensics
Computer music and recording
Computer-aided music creation/computer-aided music teaching system/computer music production technology/computer music software development/audio and multi-channel sound system/sound device and related multimedia technology/sound effect and sound design/audio human-computer interaction
Auditory psychology
Multimedia application combining hearing and vision
It is worth mentioning that last year's CSMT conference held two special sessions: one was to discuss computer hearing for general audio, trying to expand the application of audio +AI artificial intelligence in all walks of life except music, such as marine ship identification, equipment diagnosis, AI medical treatment, voice acoustics, audio monitoring, animal identification, agricultural protection, industrial automation, etc. The other was to discuss the cross-integration of China folk music and computer science and technology, which reflected the foresight of domestic conferences.
The popular AI+ music algorithm at present
For the current research of music artificial intelligence algorithm, Professor Fu Xiaodong of the Music Department of China Conservatory of Music divided it into two categories: self-discipline and heteronomy in the article "Ethical Thinking of Music Artificial Intelligence-Algorithm Composition" published in Art Exploration 2018.05.
Among them, "self-discipline" means that the machine strictly or not strictly follows the internal structure principles stipulated in advance to generate music works corresponding to audio materials, and the final audio presentation is limited by the internal structure principles; "Heteronomy" means that machines strictly follow the external structural principles stipulated according to human experience and generate works by mapping them into acoustics. The final sound presentation is limited by the heteronomy of the external structure principle.
The final combing results are as follows:
"Self-discipline" Music Artificial Intelligence Algorithm
(A) mathematical model (mathematical model)
Compose music with a mathematical model composed of mathematical algorithms and random events. Among them, the algorithm is equivalent to the law of composition, and random events are equivalent to musical elements-all elements in music can be decomposed into a series of random events, such as four attributes of sound, three elements of music and so on. Composers (programmers) give them different weights, and use a specific random algorithm to calculate and process them to get sound sequences, and the results are uncertain. Commonly used stochastic algorithms are Markov chain and Gaussian distribution. At present, the music artificial intelligence works based on mathematical model have a considerable sense of "intelligence" in the aspects of accompaniment speed, dynamic processing of phrases, termination of stretching rhythm, etc., but there is still an obvious lack in the overall audibility of the works.
(2) Evolutionary method.
Evolutionary algorithm originates from the theory of biological evolution revealed by Darwin, and uses the algorithm to simulate the process of species evolution to construct music works. Random or artificial acoustic events are grouped into a population, and the existing individuals in the population are eliminated by the iterative algorithm of seed selection, heredity and mutation, and the results are corrected by the audit program composed of fitness function to ensure the quality of its aesthetic significance. The most common evolutionary computation methods are genetic algorithm and genetic programming. The logic of evolutionary algorithm trying to match the process of species evolution with the process of music generation is not perfect, so the aesthetic recognition of works is not high, and now it is mostly used in harmony configuration and accompaniment tasks.
Grammars (grammars)
The rules of music formation can be compared with the grammatical rules of human language. Human language is composed of words, phrases and sentences according to certain grammatical rules. Motivation, festivals and phrases in music also have similar structural characteristics. Firstly, a specific grammatical rule of a musical work is created, and a musical work is finally generated by combining various musical materials such as harmony, rhythm and pitch. It is true that music and language are isomorphic to some extent, but in comparison, music rules show greater flexibility and variability. Music works produced by language algorithms with fixed grammatical rules and some variable rules have some rigid and inflexible characteristics.
Artificial intelligence algorithm of heteronomy music
(A) migration model algorithm (translation model)
Information in non-music media signal sources is mapped and migrated to music audio information. The most common is to convert visual information, such as converting lines in an image into melody, color into harmony, and chroma into intensity; The spatial displacement of a moving object is converted into melody, and the speed is converted into beat rhythm. It can also be used to convey non-visual information, such as transforming positive/negative descriptions in literary works into primary/minor chords through automatic sentiment analysis system. In fact, human senses do have a "synesthesia" effect to some extent, such as the correspondence between spatial lines and melody trends, but if they are mapped strictly, there is no strong psychological evidence. Therefore, the music works generated by the migration model algorithm often appear in interactive new media art performances, which are more relevant and interactive in aesthetic taste. However, once music works are presented separately from their mapping objects, the audibility of such works will be greatly reduced.
Knowledge-based system.
Based on the knowledge base of a certain music style, the aesthetic characteristics of this music style are extracted and coded, that is, inductive reasoning; Taking the coding program as the algorithm, that is, deductive reasoning, we can create new works with similar styles. For example, baroque music style coding based on counterpoint principle, classical romantic music style coding based on large and small harmony system, impressionist music style coding with weakened harmony function and the generation of corresponding style works all belong to knowledge reasoning system algorithms. The algorithm is close to the learning process of the theory of composition technology in the Conservatory of Music to a certain extent, and the generated music works are very similar to the specific style knowledge base on which they are based, with high audibility. Its disadvantage lies in the relative separation of induction and deduction, that is, the style code must be provided by the operator, and the program itself is only the execution of the code. The operator's abstract understanding of the law of creation will seriously affect the result of the work, and there will be shortcomings of rigidity and similarity.
Machine Learning (machine learning)
The operator inputs a lot of music stereo for the computer, and the computer effectively "listens to and learns" the rules of music composition, that is, the process is similar to the knowledge reasoning system, but the operator does not strictly specify the music type, nor does he provide style coding for the program, and this process is automatically completed by the algorithm program, emphasizing its autonomy and unsupervised learning. Of course, in essence, "unsupervised" machine learning can only be to a certain extent and scope, and it is still limited by the knowledge base provided by the operator. Machine learning is closely related to the research results of computational science, such as mathematical optimization and data mining, and to the research results of cognitive science and neural network, among which decision tree, artificial neural network and deep learning are the most obvious methods, which are the algorithms with the highest degree of imitation to the biological learning process so far. Machine learning still belongs to bionics, but it transcends the bionics of structure and mechanics and is the bionics of human brain thinking process. Machine learning can be used not only for general music creation, but also for improvisation and competition. Although music works with various specified styles or mixed styles can be generated, it still depends on the type of music data provided by the operator, and it is a regular acoustic prediction through the probability statistics of random events.
According to Professor Fu's classification standard, we will be able to effectively classify most popular artificial intelligence+music research work.
It is worth mentioning that, with the cooperation of China University of Science and Technology, Microsoft Institute of Artificial Intelligence and Suzhou University, the paper Xiaoice Band: A Melody and Arrangement Generation Framework for Popular Music, which describes the end-to-end melody generation and arrangement generation framework of songs, was successfully awarded the best student paper on research track in KDD 20 18. Lei Feng's AI Technology Review made a corresponding interpretation of this. Interested readers can click/news/2065438+0808/nkobbblrdhxzsyadg5.html to look back.
Generally speaking, artificial intelligence will play a more important role in the music field in the future. It can help people analyze, create and share a lot of repetitive works, further stimulate creativity and explore the possibilities of music form and content. I hope that through this interdisciplinary and integrated cooperation, we can sum up and improve all kinds of music creation logic, make breakthroughs in perception and emotion, make artificial intelligence innovate in many fields of music, and have an impact on teaching and social services.