Piano fingering automatic generation method based on fingering rules and reinforcement learning

Abstract

The invention discloses a piano fingering automatic generation method based on fingering rules and reinforcement learning, and the method comprises the following steps: establishing a fingering quantitative evaluation system according to piano fingering conversion rules; analyzing the MIDI file of the music and extracting a key position sequence; training by adopting a time sequence difference control reinforcement learning algorithm as a learning network; and inputting the MIDI file of the new music, and generating a fingering sequence by the learning network according to the key position sequence decision finger serial number. Based on the general fingering rules and reinforcement learning, the fingering rules descriptive of characters are used for automatically generating a fingering sequence through machine learning. The method does not need fingering data for training, so that the method is not limited by fingering data quality and size. Besides, a time sequence difference controltype reinforcement learning algorithm is adopted, an environment model, rewards and a probability distribution model of the next state do not need to be established, the method is simple and clear, avalue function is updated in real time in each step, and the learning speed is increased.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence machine learning, in particular to a method for automatically generating piano fingerings based on fingering rules and reinforcement learning. Background technique [0002] Piano education is the largest category in the music education market. At present, the proportion of piano learning in kindergartens is more than 60%, and that of primary schools is 30%. But what does not match it is that the current piano teachers are limited and the teaching level is uneven; the teaching resources lack creativity and cannot keep up with the needs of the society; the teaching methods are monotonous and not targeted. [0003] Fingering is an important part that piano learners need to master. For piano beginners, especially groups such as the elderly and the disabled, fingering exercises are more difficult. There are many problems such as difficulty in reading notation, poor coherence in playin...

AI Technical Summary

Problems solved by technology

However, this method of seeking the maximum probability fingering path based on fingering data is limited by the quality of the data itself and the size of the data, and the generated results can only be close to the fingering level of the fingering data, but cannot be surpassed.

[0006] In addition, due to the special nature of piano fingering, there will be differences in the styles of fingering rules of different genres, and the fingering styles generated based on different data are also different. It is difficult to say which is better and which is worse.

Moreover, due to the inability to guarantee the quality of fingering data and the limited amount of data, the fingerings generated through data learning may not be comprehensive or universal.

Images