A Lightweight Deep Learning-Based Approach for Jazz Music Generation in MIDI Format

Abstract

In today’s real-world, estimation of the level of difficulty of the musical is part of very meaningful musical learning. A musical learner cannot learn without a defined precise estimation. This problem is not very basic but it is complicated up to some extent because of the subjectivity of the contents and the scarcity of the data. In this paper, a lightweight model that generates original music content using deep learning along with generating music based on a specific genre is proposed. The paper discusses a lightweight deep learning-based approach for jazz music generation in MIDI format. In this work, the genre of music chosen is Jazz, and the songs selected are classical numbers composed by various artists. All the songs are in MIDI format and there might be differences in the pace or tone of the music. It is prudential to make sure that the chosen datasets that do not have these kinds of differences and are similar to the final output as desired. A model is trained to take in a part of a music file as input and should produce its continuation. The result generated should be similar to the dataset given as the input. Moreover, the proposed model also generates music using a particular instrument.

Reinforcement Learning-based Algorithms for Music Improvisation and Arrangement in Sensor Networks for the Internet of Things

“…Figure 3.1 has articulated suggested model outcomes for various datasets. The data sets are "midi files dataset 1", "midi files dataset 2", and the "classical dataset" [12]. It is done by evaluating the dataset size in minutes, the numeral of ages vary, the period for each age in seconds, set size, initial losing value, and last losing worth.…”

“…It has been found that "reinforcement learning-based algorithms" incorporated with sensor networks propose compelling opportunities for improving "music improvisation" and interpretation. Intelligent agents can comprehend and adapt their musical determination, resulting in more graphic and interactive music concerts in the IoT era, by leveraging real-time data from detectors [12].…”

“…Valid sensor arrangement, data accompaniment, and bluster removal processes are essential for sufficient music data assembling. datasets are "midi files dataset 1", "midi files dataset 2", and the "classical dataset" [12]. It is obvious from the picture that as the numeral of age is improved the failure worth declines.…”

“…The failure worth at the 10 age is 4.12 for "midi files dataset 1", 4.56 for "midi files dataset 2", and 2.60 for the "classical dataset". In the middle of the age when the worth of these datasets is 50, the loss value is 0.39 for "midi files dataset 1", 0.34 for "midi files dataset 2", and 0.11 for the "classical dataset" [12]. The failure worth at the 100 age is 0.03 for "midi files dataset 1", 0.06 for "midi files dataset 2", and 0.01 for the "classical dataset".…”

Multi-Genre Symbolic Music Generation using Deep Convolutional Generative Adversarial Network

“…It is worth noting that previous research on MIDI music generation has primarily focused on well-known genres such as classical, jazz, and hip-hop [2,3]. As a result, the potential of MIDI music generation for creating music across a broader range of genres has not been fully explored or realized.…”