Composition of music using human evaluation

Unehara, Muneyuki; Onisawa, Takehisa

doi:10.1109/fuzz.2001.1008872

Cited by 12 publications

(4 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A more exotic solution was employed by previously cited Tokui and Iba (2000): training a neural network to filter out candidates with low fitness, thus presenting to the user individuals of acceptable quality. Unehara and Onisawa (2001) presented just 10% of candidate melodies to the human user, and then parts of the genomes of the best rated ones were dispersed in the population by "horizontal gene transfer". McDermott and O'Reilly (2011) also limited the number of candidates exposed to user rating, filtering out the worst ones with heuristic functions.…”

Section: Musical Igasmentioning

confidence: 99%

AI Methods in Algorithmic Composition: A Comprehensive Survey

Fernández

Vico

2013

jair

187

107

View full text Add to dashboard Cite

Algorithmic composition is the partial or total automation of the process of music composition by using computers. Since the 1950s, different computational techniques related to Artificial Intelligence have been used for algorithmic composition, including grammatical representations, probabilistic methods, neural networks, symbolic rule-based systems, constraint programming and evolutionary algorithms. This survey aims to be a comprehensive account of research on algorithmic composition, presenting a thorough view of the field for researchers in Artificial Intelligence.

show abstract

Section: Musical Igasmentioning

confidence: 99%

AI Methods in Algorithmic Composition: A Comprehensive Survey

Fernández

Vico

2013

jair

187

107

View full text Add to dashboard Cite

show abstract

“…Our project of SBEAT is now still less than 2 years old. Some researchers are struggling to build similar systems using their original ideas of musical application of IEC technique like Tokui and Iba (2000) and Unehara and Onisawa (2001). Including the previous pioneers' approaches, it may be fruitful to combine several different ideas to build up more practical integrated system to support human activities of music composition.…”

Section: Future Workmentioning

confidence: 99%

Simulated breeding – a framework of breeding artifacts on the computer

Unemi

2003

Kybernetes

View full text Add to dashboard Cite

A basic framework of Simulated Breeding, a type of interactive evolutionary computing to breed artifacts of which origin is Blind Watchmaker by R. Dawkins is considered. These methods, make it easy for humans; to design a complex object adapted to his/her subjective criteria, just similar to agricultural products that we have been developing over thousands of years. Starting from randomly initialised genome, the solution candidates are improved through several generations with artificial selection. Graphical user interface helps the process of breeding with techniques of multi‐field user interface and partial breeding. The former improves the diversity of individuals that prevents being trapped at local optimum. The latter makes it possible for the user to fix features that he/she already satisfied. These methods were examined through artistic applications by the author, SBART for graphics art and SBEAT for music. Combining with direct genome editor and exportation to another graphical or musical tool on the computer, they can be powerful tools for artistic creation. These systems may contribute to create a type of new culture.

show abstract

“…However, few objective algorithms or indicators has been put forward to assess whether a melody was created by a machine or a person. Although several attempts has been made, such as measures from information theory to compare Bach's music [5], or probability transfer relation with the N -gram model to compare British and American folk music melody [1], most of the distinguishing task on composing works are made in human evaluation, namely let some people listen to the music produced by the AI or human, and evaluate its similarity with music made of man or AI [6][7][8][9][10].…”

Section: Introductionmentioning

confidence: 99%

A Transformer Based Pitch Sequence Autoencoder with MIDI Augmentation

Ding

2020

Preprint

View full text Add to dashboard Cite

Algorithms based on deep learning have been widely put forward for automatic music generated. However, few objective approaches have been proposed to assess whether a melody was created by automatons or Homo sapiens. Conference of Sound and Music Technology (2020) provides us a great opportunity to cope with the problem. In this paper, a masked language model based on ALBERT trained with AI-composed single-track MIDI is demonstrated for composers classification tasks. Besides, music tune transposition and MIDI sequence truncation is applied for data augments. To prevent from over-fitting, a refined loss function is proposed and the amount of parameters is reduced. This work provides a new approach to tackle the problem on obtaining features from tiny dataset which is common in music signal analysis and deserve more attention.

show abstract

Composition of music using human evaluation

Cited by 12 publications

References 1 publication

AI Methods in Algorithmic Composition: A Comprehensive Survey

AI Methods in Algorithmic Composition: A Comprehensive Survey

Simulated breeding – a framework of breeding artifacts on the computer

A Transformer Based Pitch Sequence Autoencoder with MIDI Augmentation

Contact Info

Product

Resources

About